ItemPick and BoxPick

Introduction

The ItemPick and BoxPick modules provide out-of-the-box perception solutions for robotic pick-and-place applications. ItemPick targets the detection of flat surfaces of unknown objects for picking with a suction gripper. BoxPick detects rectangular surfaces and determines their position, orientation and size for grasping. With the +Match extension, BoxPick can be used to detect textured rectangles with consistent orientations. The interface of both modules is very similar. Therefore both modules are described together in this chapter.

In addition, both modules offer:

  • A dedicated page on the rc_visard Web GUI for easy setup, configuration, testing, and application tuning.
  • The definition of regions of interest to select relevant volumes in the scene (see RoiDB).
  • A load carrier detection functionality for bin-picking applications (see LoadCarrier), to provide grasps for items inside a bin only.
  • The definition of compartments inside a load carrier to provide grasps for specific volumes of the bin only.
  • Support for static and robot-mounted cameras and optional integration with the Hand-eye calibration module, to provide grasps in the user-configured external reference frame.
  • A quality value associated to each suggested grasp and related to the flatness of the grasping surface.
  • Selection of a sorting strategy to sort the returned grasps.
  • 3D visualization of the detection results with grasp points and gripper animations in the Web GUI.

Note

In this chapter, cluster and surface are used as synonyms and identify a set of points (or pixels) with defined geometrical properties.

The modules are optional on-board modules of the rc_visard and require separate ItemPick or BoxPick licenses to be purchased. The +Match extension of BoxPick requires an extra license.

Detection of items (BoxPick)

There are two different types of models for the rectangles to be detected by the BoxPick module.

Per default, BoxPick only supports item_models of type RECTANGLE. With the +Match extension, also item models of type TEXTURED_BOX can be detected. The detection of the different item model types is described below.

Optionally, further information can be given to the BoxPick module:

  • The ID of the load carrier which contains the items to be detected.
  • A compartment inside the load carrier where to detect items.
  • The ID of the region of interest where to search for the load carriers if a load carrier is set. Otherwise, the ID of the region of interest where to search for the items.
  • The current robot pose in case the camera is mounted on the robot and the chosen coordinate frame for the poses is external or the chosen region of interest is defined in the external frame.

The returned pose of a detected item is the pose of the center of the detected rectangle in the desired reference frame (pose_frame), with its z axis pointing towards the camera and the x axis aligned with the long side of the item. This pose has a 180° rotation ambiguity around the z axis, which can be resolved by using the +Match extension with a TEXTURED_BOX item model. Each detected item includes a uuid (Universally Unique Identifier) and the timestamp of the oldest image that was used to detect it.

Detection of items of type RECTANGLE (BoxPick)

BoxPick supports multiple item_models of type RECTANGLE. Each item model is defined by its minimum and maximum size, with the minimum dimensions strictly smaller than the maximum dimensions. The dimensions should be given fairly accurately to avoid misdetections, while still considering a certain tolerance to account for possible production variations and measurement inaccuracies.

The detection of the rectangles runs in several steps. First, the point cloud is segmented into preferably plane clusters. Then, straight line segments are detected in the 2D images and projected onto the corresponding clusters. The clusters and the detected lines are visualized in the “Intermediate Result” visualization on the Web GUI’s BoxPick page. Finally, for each cluster, the set of rectangles best fitting to the detected line segments is extracted.

Detection of items of type TEXTURED_BOX (BoxPick+Match)

With the +Match extension, BoxPick additionally supports item_models of type TEXTURED_BOX. When this item model type is used, only one item model can be given for each request.

The TEXTURED_BOX item model type should be used to detect multiple rectangles that have the same texture, i.e. the same look or print, such as printed product packaging, labels, brochures or books. It is required that for all objects the texture is at the same position with respect to the object geometry. Furthermore, the texture should not be repetitive.

A TEXTURED_BOX item is defined by the item’s exact dimensions x, y and z (currently z must always be 0) with a tolerance dimensions_tolerance_m that indicates, how much the detected dimensions are allowed to deviate from the given dimensions. By default, a tolerance of 0.01 m is assumed. Furthermore, a template_id must be given, which will be used to refer to the specified dimensions and the textures of the detected rectangles. Additionally, the maximum possible deformation of the items max_deformation_m can be given in meters (default 0.004 m), to account for rigid or more flexible objects.

If a template_id is used for the first time, BoxPick will run the detection of rectangles as for the item model type RECTANGLE, and use the given dimensions tolerance to specify the dimensions range. From the detected rectangles, so-called views are created, which contain the shape and the image intensity values of the rectangles, and are stored in a newly created template with the given template_id. The views are created iteratively: Starting from the detected rectangle with the highest score, a view is created and then used to detect more rectangles with the same texture. Then, all remaining clusters are used to detect further rectangles by the given dimensions range and again a view is created from the best rectangle and used for further detections. Each template can store up to 10 different views, for example corresponding to different types of the same product packaging. Each view will be assigned a unique ID (view_uuid) and all rectangle items with a matching texture will be assigned the same view_uuid. That also means that all items with the same view_uuid will have consistent orientations, because the orientation of each item is aligned with its texture. The views can be displayed, deleted and the orientation of each view can be set via the Web GUI by clicking on the template or its edit symbol in the template list. Each detected item contains a field view_pose_set indicating whether the orientation of the item’s view was explicitly set or is still unset at its original random state, which has a 180° ambiguity. The type of a returned item with a view_uuid will be TEXTURED_RECTANGLE.

If the template with the given template_id already exists, the existing views will be used to detect rectangles based on their texture. If additional rectangles are found with matching dimensions, but different texture, new views will be generated and added to the template. When the maximum number of views is reached, views that are matched only rarely will be deleted so that newly generated views can be added to the template and the template is kept up-to-date. To prevent a template from being updated, automatic updating can be disabled and enabled for each template in the Web GUI by clicking on the template or the edit symbol in the template list. The dimension tolerance and the maximum deformation can also be changed there for each template. The maximum deformation determines the tolerance for the texture matching, representing possible shifts within the texture, e.g. caused by deformations of the object surface. For rigid objects the max_deformation_m should be set to a low value in meters to ensure accurate matching.

The template’s dimensions can only be specified when creating a new template. Once the template is generated, the dimensions cannot be changed and do not need to be given in the detect request. If the dimensions are still given in the request, they must match the existing dimensions in the template. However, the dimensions_tolerance_m and max_deformation_m can be set differently in every detect request and their values will also be updated in the stored template.

Computation of grasps

The ItemPick and BoxPick modules offer a service for computing grasps for suction grippers. The gripper is defined by its suction surface length and width.

The ItemPick module identifies flat surfaces in the scene and supports flexible and/or deformable items. The type of these item_models is called UNKNOWN since they don’t need to have a standard geometrical shape. Optionally, the user can also specify the minimum and maximum size of the item.

For BoxPick, the grasps are computed on the detected rectangular items (see Detection of items (BoxPick)).

Optionally, further information can be given to the modules in a grasp computation request:

  • The ID of the load carrier which contains the items to be grasped.
  • A compartment inside the load carrier where to compute grasps (see Load carrier compartments).
  • The ID of the 3D region of interest where to search for the load carriers if a load carrier is set. Otherwise, the ID of the 3D region of interest where to compute grasps.
  • Collision detection information: The ID of the gripper to enable collision checking and optionally a pre-grasp offset to define a pre-grasp position. Details on collision checking are given below in CollisionCheck.

A grasp provided by the ItemPick and BoxPick modules represents the recommended pose of the TCP (Tool Center Point) of the suction gripper. The grasp type is always set to SUCTION. The computed grasp pose is the center of the biggest ellipse that can be inscribed in each surface. The grasp orientation is a right-handed coordinate system and is defined such that its z axis is normal to the surface pointing inside the object at the grasp position and its x axis is directed along the maximum elongation of the ellipse.

_images/itempick_grasp_surface.png

Fig. 32 Illustration of suction grasp with coordinate system and ellipse representing the maximum suction surface.

Each grasp includes the dimensions of the maximum suction surface available, modelled as an ellipse of axes max_suction_surface_length and max_suction_surface_width. The user is enabled to filter grasps by specifying the minimum suction surface required by the suction device in use.

In the BoxPick module, the grasp position corresponds to the center of the detected rectangle and the dimensions of the maximum suction surface available matches the estimated rectangle dimensions. Detected rectangles with missing data or occlusions by other objects for more than 15% of their surface do not get an associated grasp.

Each grasp also includes a quality value, which gives an indication of the flatness of the grasping surface. The quality value varies between 0 and 1, where higher numbers correspond to a flatter reconstructed surface.

The grasp definition is complemented by a uuid (Universally Unique Identifier) and the timestamp of the oldest image that was used to compute the grasp.

Grasp sorting is performed based on the selected sorting strategy. The following sorting strategies are available and can be set in the Web GUI or using the set_sorting_strategies service call:

  • gravity: highest grasp points along the gravity direction are returned first,
  • surface_area: grasp points with the largest surface area are returned first,
  • direction: grasp points with the shortest distance along a defined direction vector in a given pose_frame are returned first.

If no sorting strategy is set or default sorting is chosen in the Web GUI, sorting is done based on a combination of gravity and surface_area.

Interaction with other modules

Internally, the ItemPick and BoxPick modules depend on, and interact with other on-board modules as listed below.

Note

All changes and configuration updates to these modules will affect the performance of the ItemPick and BoxPick modules.

Stereo camera and Stereo matching

The ItemPick and BoxPick modules make internally use of the following data:

  • Rectified images from the Camera module (rc_camera);
  • Disparity, error, and confidence images from the Stereo matching module (rc_stereomatching).

All processed images are guaranteed to be captured after the module trigger time.

IO and Projector Control

In case the rc_visard is used in conjunction with an external random dot projector and the IO and Projector Control module (rc_iocontrol), it is recommended to connect the projector to GPIO Out 1 and set the stereo-camera module’s acquisition mode to SingleFrameOut1 (see Stereo matching parameters), so that on each image acquisition trigger an image with and without projector pattern is acquired.

Alternatively, the output mode for the GPIO output in use should be set to ExposureAlternateActive (see Description of run-time parameters).

In either case, the Auto Exposure Mode exp_auto_mode should be set to AdaptiveOut1 to optimize the exposure of both images (see Stereo camera parameters).

Hand-eye calibration

In case the camera has been calibrated to a robot, the ItemPick and BoxPick modules can automatically provide poses in the robot coordinate frame. For the ItemPick and BoxPick nodes’ Services, the frame of the output poses can be controlled with the pose_frame argument.

Two different pose_frame values can be chosen:

  1. Camera frame (camera). All poses provided by the modules are in the camera frame, and no prior knowledge about the pose of the camera in the environment is required. This means that the configured regions of interest and load carriers move with the camera. It is the user’s responsibility to update the configured poses if the camera frame moves (e.g. with a robot-mounted camera).
  2. External frame (external). All poses provided by the modules are in the external frame, configured by the user during the hand-eye calibration process. The module relies on the on-board Hand-eye calibration module to retrieve the sensor mounting (static or robot mounted) and the hand-eye transformation. If the mounting is static, no further information is needed. If the sensor is robot-mounted, the robot_pose is required to transform poses to and from the external frame.

Note

If no hand-eye calibration is available, all pose_frame values should be set to camera.

All pose_frame values that are not camera or external are rejected.

If the sensor is robot-mounted, the current robot_pose has to be provided depending on the value of pose_frame and the definition of the sorting direction:

  • If pose_frame is set to external, providing the robot pose is obligatory.
  • If the sorting direction is defined in external, providing the robot pose is obligatory.
  • In all other cases, providing the robot pose is optional.

LoadCarrier

The ItemPick and BoxPick modules use the load carrier detection functionality provided by the LoadCarrier module (rc_load_carrier), with the run-time parameters specified for this module. However, only one load carrier will be returned and used in case multiple matching load carriers could be found in the scene. In case multiple load carriers of the same type are visible, a 3D region of interest should be set to ensure that always the same load carrier is used for the ItemPick and BoxPick modules.

CollisionCheck

Collision checking can be easily enabled for grasp computation of the ItemPick and BoxPick modules by passing the ID of the used gripper and optionally a pre-grasp offset to the compute_grasps service call. The gripper has to be defined in the GripperDB module (see Setting a gripper) and details about collision checking are given in Collision checking within other modules.

If collision checking is enabled, only grasps which are collision free will be returned. However, the visualization images on the ItemPick or BoxPick page of the Web GUI also show colliding grasp points as black ellipses.

The CollisionCheck module’s run-time parameters affect the collision detection as described in CollisionCheck Parameters.

Parameters

The ItemPick and BoxPick modules are called rc_itempick and rc_boxpick in the REST-API and are represented in the Web GUI under Modules ‣ ItemPick and Modules ‣ BoxPick. The user can explore and configure the rc_itempick and rc_boxpick module’s run-time parameters, e.g. for development and testing, using the Web GUI or the REST-API interface.

Parameter overview

These modules offer the following run-time parameters:

Table 33 The rc_itempick and rc_boxpick modules’ application parameters
Name Type Min Max Default Description
max_grasps int32 1 20 5 Maximum number of provided grasps
Table 34 The rc_itempick and rc_boxpick modules’ surface clustering parameters
Name Type Min Max Default Description
cluster_max_dimension float64 0.05 0.8 0.3 Only for rc_itempick. Maximum allowed diameter for a cluster in meters. Clusters with a diameter larger than this value are not used for grasp computation.
cluster_max_curvature float64 0.005 0.5 0.11 Maximum curvature allowed within one cluster. The smaller this value, the more clusters will be split apart.
clustering_patch_size int32 3 10 4 Only for rc_itempick. Size in pixels of the square patches the depth map is subdivided into during the first clustering step
clustering_max_surface_rmse float64 0.0005 0.01 0.004 Maximum root-mean-square error (RMSE) in meters of points belonging to a surface
clustering_discontinuity_factor float64 0.1 5.0 1.0 Factor used to discriminate depth discontinuities within a patch. The smaller this value, the more clusters will be split apart.
Table 35 The rc_boxpick module’s rectangle detection parameters
Name Type Min Max Default Description
mode string - - Unconstrained Mode of the rectangle detection: [Unconstrained, PackedGridLayout, PackedLayers]
manual_line_sensitivity bool false true false Indicates whether the user-defined line sensitivity should be used or the automatic one
line_sensitivity float64 0.1 1.0 0.1 Sensitivity of the line detector
prefer_splits bool false true false Indicates whether rectangles are split into smaller ones when possible
min_cluster_coverage float64 0.0 0.99 0.0 Gives the minimal ratio of points per cluster that must be covered with detected items.

Description of run-time parameters

Each run-time parameter is represented by a row on the Web GUI’s ItemPick or BoxPick page. The name in the Web GUI is given in brackets behind the parameter name and the parameters are listed in the order they appear in the Web GUI:

max_grasps (Maximum Grasps)

sets the maximum number of provided grasps.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/parameters/parameters?max_grasps=<value>
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/parameters?max_grasps=<value>

cluster_max_dimension (Only for ItemPick, Cluster Maximum Dimension)

is the maximum allowed diameter for a cluster in meters. Clusters with a diameter larger than this value are not used for grasp computation.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_itempick/parameters/parameters?cluster_max_dimension=<value>
PUT http://<host>/api/v1/nodes/rc_itempick/parameters?cluster_max_dimension=<value>

cluster_max_curvature (Cluster Maximum Curvature)

is the maximum curvature allowed within one cluster. The smaller this value, the more clusters will be split apart.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/parameters/parameters?cluster_max_curvature=<value>
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/parameters?cluster_max_curvature=<value>

clustering_patch_size (Only for ItemPick, Patch Size)

is the size of the square patches the depth map is subdivided into during the first clustering step in pixels.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_itempick/parameters/parameters?clustering_patch_size=<value>
PUT http://<host>/api/v1/nodes/rc_itempick/parameters?clustering_patch_size=<value>

clustering_discontinuity_factor (Discontinuity Factor)

is the factor used to discriminate depth discontinuities within a patch. The smaller this value, the more clusters will be split apart.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/parameters/parameters?clustering_discontinuity_factor=<value>
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/parameters?clustering_discontinuity_factor=<value>

clustering_max_surface_rmse (Maximum Surface RMSE)

is the maximum root-mean-square error (RMSE) in meters of points belonging to a surface.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/parameters/parameters?clustering_max_surface_rmse=<value>
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/parameters?clustering_max_surface_rmse=<value>

mode (Only for BoxPick, Mode)

determines the mode of the rectangle detection. Possible values are Unconstrained, PackedGridLayout and PackedLayers. In PackedGridLayout mode, rectangles of a cluster are detected in a dense grid pattern. In PackedLayers mode, boxes are assumed to form layers and box detection will start searching for items at the cluster corners. Use this mode in de-palletizing applications. In Unconstrained mode (default), rectangles are detected without posing any constraints on their relative locations or their positions in the segmented cluster. Fig. 33 illustrates the modes for different scenarios.

_images/boxpick_modes_new.svg

Fig. 33 Illustration of appropriate BoxPick modes for different scenes. Modes marked with yellow are applicable but not recommended for the corresponding scene. The gray areas indicate the rectangles to be detected.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/parameters/parameters?mode=<value>
PUT http://<host>/api/v1/nodes/rc_boxpick/parameters?mode=<value>

manual_line_sensitivity (Only for BoxPick, Manual Line Sensitivity)

determines whether the user-defined line sensitivity should be used to extract the lines for rectangle detection. If this parameter is set to true, the user-defined line_sensitivity value will be used. If this parameter is set to false, automatic line sensitivity will be used. This parameter should be set to true when automatic line sensitivity does not give enough lines at the box boundaries so that boxes cannot be detected. The detected line segments are visualized in the “Intermediate Result” visualization on the Web GUI’s BoxPick page.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/parameters/parameters?manual_line_sensitivity=<value>
PUT http://<host>/api/v1/nodes/rc_boxpick/parameters?manual_line_sensitivity=<value>

line_sensitivity (Only for BoxPick, Line Sensitivity)

determines the line sensitivity for extracting the lines for rectangle detection, if the parameter manual_line_sensitivity is set to true. Otherwise, the value of this parameter has no effect on the rectangle detection. Higher values give more line segments, but also increase the runtime of the box detection. This parameter should be increased when boxes cannot be detected because their boundary edges are not detected. The detected line segments are visualized in the “Intermediate Result” visualization on the Web GUI’s BoxPick page.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/parameters/parameters?line_sensitivity=<value>
PUT http://<host>/api/v1/nodes/rc_boxpick/parameters?line_sensitivity=<value>

prefer_splits (Only for BoxPick, Prefer Splits)

determines whether rectangles should be split into smaller ones if the smaller ones also match the given item models. This parameter should be set to true for packed box layouts in which the given item models would also match a rectangle of the size of two adjoining boxes. If this parameter is set to false, the larger rectangles will be preferred in these cases.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/parameters/parameters?prefer_splits=<value>
PUT http://<host>/api/v1/nodes/rc_boxpick/parameters?prefer_splits=<value>

min_cluster_coverage (Only for BoxPick, Minimum Cluster Coverage)

determines which ratio of each segmented cluster must be covered with rectangle detections to consider the detections to be valid. If the minimum cluster coverage is not reached for a cluster, no rectangle detections will be returned for this cluster and a warning will be given. This parameter should be used to verify that all items on a layer in a de-palletizing scenario are detected.

Via the REST-API, this parameter can be set as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/parameters/parameters?min_cluster_coverage=<value>
PUT http://<host>/api/v1/nodes/rc_boxpick/parameters?min_cluster_coverage=<value>

Status values

The rc_itempick and rc_boxpick modules report the following status values:

Table 36 The rc_itempick and rc_boxpick modules status values
Name Description
data_acquisition_time Time in seconds required by the last active service to acquire images
grasp_computation_time Processing time of the last grasp computation in seconds
last_timestamp_processed The timestamp of the last processed dataset
load_carrier_detection_time Processing time of the last load carrier detection in seconds
processing_time Processing time of the last detection (including load carrier detection) in seconds
state The current state of the rc_itempick and rc_boxpick node

The reported state can take one of the following values.

Table 37 Possible states of the ItemPick and BoxPick modules
State name Description
IDLE The module is idle.
RUNNING The module is running and ready for load carrier detection and grasp computation.
FATAL A fatal error has occurred.

Services

The user can explore and call the rc_itempick and rc_boxpick module’s services, e.g. for development and testing, using the REST-API interface or the rc_visard Web GUI.

The ItemPick and BoxPick modules offer the following services.

detect_items (BoxPick only)

Triggers the detection of rectangles as described in Detection of items (BoxPick).

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/services/detect_items
PUT http://<host>/api/v1/nodes/rc_boxpick/services/detect_items

Required arguments:

pose_frame: see Hand-eye calibration.

item_models: list of item models to be detected. The type of the item model must be RECTANGLE or TEXTURED_BOX. For type RECTANGLE, rectangle must be filled, while for TEXTURED_BOX, textured_box must be filled. See Detection of items (BoxPick) for a detailed description of the item model types.

Potentially required arguments:

robot_pose: see Hand-eye calibration.

Optional arguments:

load_carrier_id: ID of the load carrier which contains the items to be detected.

load_carrier_compartment: compartment inside the load carrier where to detect items (see Load carrier compartments).

region_of_interest_id: if load_carrier_id is set, ID of the 3D region of interest where to search for the load carriers. Otherwise, ID of the 3D region of interest where to search for the items.

The definition for the request arguments with corresponding datatypes is:

{
  "args": {
    "item_models": [
      {
        "rectangle": {
          "max_dimensions": {
            "x": "float64",
            "y": "float64"
          },
          "min_dimensions": {
            "x": "float64",
            "y": "float64"
          }
        },
        "textured_box": {
          "dimensions": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "dimensions_tolerance_m": "float64",
          "max_deformation_m": "float64",
          "template_id": "string"
        },
        "type": "string"
      }
    ],
    "load_carrier_compartment": {
      "box": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "pose": {
        "orientation": {
          "w": "float64",
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "position": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        }
      }
    },
    "load_carrier_id": "string",
    "pose_frame": "string",
    "region_of_interest_id": "string",
    "robot_pose": {
      "orientation": {
        "w": "float64",
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "position": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      }
    }
  }
}

load_carriers: list of detected load carriers.

items: list of detected rectangles.

timestamp: timestamp of the image set the detection ran on.

return_code: holds possible warnings or error codes and messages.

The definition for the response with corresponding datatypes is:

{
  "name": "detect_items",
  "response": {
    "items": [
      {
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "rectangle": {
          "x": "float64",
          "y": "float64"
        },
        "template_id": "string",
        "timestamp": {
          "nsec": "int32",
          "sec": "int32"
        },
        "type": "string",
        "uuid": "string",
        "view_pose_set": "bool",
        "view_uuid": "string"
      }
    ],
    "load_carriers": [
      {
        "height_open_side": "float64",
        "id": "string",
        "inner_dimensions": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "outer_dimensions": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "overfilled": "bool",
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "rim_ledge": {
          "x": "float64",
          "y": "float64"
        },
        "rim_step_height": "float64",
        "rim_thickness": {
          "x": "float64",
          "y": "float64"
        },
        "type": "string"
      }
    ],
    "return_code": {
      "message": "string",
      "value": "int16"
    },
    "timestamp": {
      "nsec": "int32",
      "sec": "int32"
    }
  }
}

compute_grasps (for ItemPick)

Triggers the computation of grasping poses for a suction device as described in Computation of grasps.

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_itempick/services/compute_grasps
PUT http://<host>/api/v1/nodes/rc_itempick/services/compute_grasps

Required arguments:

pose_frame: see Hand-eye calibration.

suction_surface_length: length of the suction device grasping surface.

suction_surface_width: width of the suction device grasping surface.

Potentially required arguments:

robot_pose: see Hand-eye calibration.

Optional arguments:

load_carrier_id: ID of the load carrier which contains the items to be grasped.

load_carrier_compartment: compartment inside the load carrier where to compute grasps (see Load carrier compartments).

region_of_interest_id: if load_carrier_id is set, ID of the 3D region of interest where to search for the load carriers. Otherwise, ID of the 3D region of interest where to compute grasps.

item_models: list of unknown items with minimum and maximum dimensions, with the minimum dimensions strictly smaller than the maximum dimensions. Only one item_model of type UNKNOWN is currently supported.

collision_detection: see Collision checking within other modules.

The definition for the request arguments with corresponding datatypes is:

{
  "args": {
    "collision_detection": {
      "gripper_id": "string",
      "pre_grasp_offset": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      }
    },
    "item_models": [
      {
        "type": "string",
        "unknown": {
          "max_dimensions": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "min_dimensions": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        }
      }
    ],
    "load_carrier_compartment": {
      "box": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "pose": {
        "orientation": {
          "w": "float64",
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "position": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        }
      }
    },
    "load_carrier_id": "string",
    "pose_frame": "string",
    "region_of_interest_id": "string",
    "robot_pose": {
      "orientation": {
        "w": "float64",
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "position": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      }
    },
    "suction_surface_length": "float64",
    "suction_surface_width": "float64"
  }
}

load_carriers: list of detected load carriers.

grasps: sorted list of suction grasps.

timestamp: timestamp of the image set the detection ran on.

return_code: holds possible warnings or error codes and messages.

The definition for the response with corresponding datatypes is:

{
  "name": "compute_grasps",
  "response": {
    "grasps": [
      {
        "item_uuid": "string",
        "max_suction_surface_length": "float64",
        "max_suction_surface_width": "float64",
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "quality": "float64",
        "timestamp": {
          "nsec": "int32",
          "sec": "int32"
        },
        "type": "string",
        "uuid": "string"
      }
    ],
    "load_carriers": [
      {
        "height_open_side": "float64",
        "id": "string",
        "inner_dimensions": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "outer_dimensions": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "overfilled": "bool",
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "rim_ledge": {
          "x": "float64",
          "y": "float64"
        },
        "rim_step_height": "float64",
        "rim_thickness": {
          "x": "float64",
          "y": "float64"
        },
        "type": "string"
      }
    ],
    "return_code": {
      "message": "string",
      "value": "int16"
    },
    "timestamp": {
      "nsec": "int32",
      "sec": "int32"
    }
  }
}

compute_grasps (for BoxPick)

Triggers the detection of rectangles and the computation of grasping poses for the detected rectangles as described in Computation of grasps.

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/rc_boxpick/services/compute_grasps
PUT http://<host>/api/v1/nodes/rc_boxpick/services/compute_grasps

Required arguments:

pose_frame: see Hand-eye calibration.

item_models: list of item models to be detected. The type of the item model must be RECTANGLE or TEXTURED_BOX. For type RECTANGLE, rectangle must be filled, while for TEXTURED_BOX, textured_box must be filled. See Detection of items (BoxPick) for a detailed description of the item model types.

suction_surface_length: length of the suction device grasping surface.

suction_surface_width: width of the suction device grasping surface.

Potentially required arguments:

robot_pose: see Hand-eye calibration.

Optional arguments:

load_carrier_id: ID of the load carrier which contains the items to be grasped.

load_carrier_compartment: compartment inside the load carrier where to compute grasps (see Load carrier compartments).

region_of_interest_id: if load_carrier_id is set, ID of the 3D region of interest where to search for the load carriers. Otherwise, ID of the 3D region of interest where to compute grasps.

collision_detection: see Collision checking within other modules.

The definition for the request arguments with corresponding datatypes is:

{
  "args": {
    "collision_detection": {
      "gripper_id": "string",
      "pre_grasp_offset": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      }
    },
    "item_models": [
      {
        "rectangle": {
          "max_dimensions": {
            "x": "float64",
            "y": "float64"
          },
          "min_dimensions": {
            "x": "float64",
            "y": "float64"
          }
        },
        "textured_box": {
          "dimensions": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "dimensions_tolerance_m": "float64",
          "max_deformation_m": "float64",
          "template_id": "string"
        },
        "type": "string"
      }
    ],
    "load_carrier_compartment": {
      "box": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "pose": {
        "orientation": {
          "w": "float64",
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "position": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        }
      }
    },
    "load_carrier_id": "string",
    "pose_frame": "string",
    "region_of_interest_id": "string",
    "robot_pose": {
      "orientation": {
        "w": "float64",
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "position": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      }
    },
    "suction_surface_length": "float64",
    "suction_surface_width": "float64"
  }
}

load_carriers: list of detected load carriers.

items: list of detected rectangles.

grasps: sorted list of suction grasps.

timestamp: timestamp of the image set the detection ran on.

return_code: holds possible warnings or error codes and messages.

The definition for the response with corresponding datatypes is:

{
  "name": "compute_grasps",
  "response": {
    "grasps": [
      {
        "item_uuid": "string",
        "max_suction_surface_length": "float64",
        "max_suction_surface_width": "float64",
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "quality": "float64",
        "timestamp": {
          "nsec": "int32",
          "sec": "int32"
        },
        "type": "string",
        "uuid": "string"
      }
    ],
    "items": [
      {
        "grasp_uuids": [
          "string"
        ],
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "rectangle": {
          "x": "float64",
          "y": "float64"
        },
        "template_id": "string",
        "timestamp": {
          "nsec": "int32",
          "sec": "int32"
        },
        "type": "string",
        "uuid": "string",
        "view_pose_set": "bool",
        "view_uuid": "string"
      }
    ],
    "load_carriers": [
      {
        "height_open_side": "float64",
        "id": "string",
        "inner_dimensions": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "outer_dimensions": {
          "x": "float64",
          "y": "float64",
          "z": "float64"
        },
        "overfilled": "bool",
        "pose": {
          "orientation": {
            "w": "float64",
            "x": "float64",
            "y": "float64",
            "z": "float64"
          },
          "position": {
            "x": "float64",
            "y": "float64",
            "z": "float64"
          }
        },
        "pose_frame": "string",
        "rim_ledge": {
          "x": "float64",
          "y": "float64"
        },
        "rim_step_height": "float64",
        "rim_thickness": {
          "x": "float64",
          "y": "float64"
        },
        "type": "string"
      }
    ],
    "return_code": {
      "message": "string",
      "value": "int16"
    },
    "timestamp": {
      "nsec": "int32",
      "sec": "int32"
    }
  }
}

set_sorting_strategies

Persistently stores the sorting strategy for sorting the grasps returned by the compute_grasps service (see Computation of grasps).

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/services/set_sorting_strategies
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/set_sorting_strategies

Only one strategy may have a weight greater than 0. If all weight values are set to 0, the module will use the default sorting strategy.

If the weight for direction is set, the vector must contain the direction vector and pose_frame must be either camera or external.

The definition for the request arguments with corresponding datatypes is:

{
  "args": {
    "direction": {
      "pose_frame": "string",
      "vector": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "weight": "float64"
    },
    "gravity": {
      "weight": "float64"
    },
    "surface_area": {
      "weight": "float64"
    }
  }
}

The definition for the response with corresponding datatypes is:

{
  "name": "set_sorting_strategies",
  "response": {
    "return_code": {
      "message": "string",
      "value": "int16"
    }
  }
}

get_sorting_strategies

Returns the sorting strategy for sorting the grasps returned by the compute-grasps service (see Computation of grasps).

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/services/get_sorting_strategies
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/get_sorting_strategies
This service has no arguments.

All weight values are 0 when the module uses the default sorting strategy.

The definition for the response with corresponding datatypes is:

{
  "name": "get_sorting_strategies",
  "response": {
    "direction": {
      "pose_frame": "string",
      "vector": {
        "x": "float64",
        "y": "float64",
        "z": "float64"
      },
      "weight": "float64"
    },
    "gravity": {
      "weight": "float64"
    },
    "return_code": {
      "message": "string",
      "value": "int16"
    },
    "surface_area": {
      "weight": "float64"
    }
  }
}

start

Starts the module. If the command is accepted, the module moves to state RUNNING.

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/services/start
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/start
This service has no arguments.

The current_state value in the service response may differ from RUNNING if the state transition is still in process when the service returns.

The definition for the response with corresponding datatypes is:

{
  "name": "start",
  "response": {
    "accepted": "bool",
    "current_state": "string"
  }
}

stop

Stops the module. If the command is accepted, the module moves to state IDLE.

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/services/stop
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/stop
This service has no arguments.

The current_state value in the service response may differ from IDLE if the state transition is still in process when the service returns.

The definition for the response with corresponding datatypes is:

{
  "name": "stop",
  "response": {
    "accepted": "bool",
    "current_state": "string"
  }
}

reset_defaults

Resets all parameters of the module to its default values, as listed in above table. Also resets sorting strategies.

Details

This service can be called as follows.

PUT http://<host>/api/v2/pipelines/0/nodes/<rc_itempick|rc_boxpick>/services/reset_defaults
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/reset_defaults
This service has no arguments.

The definition for the response with corresponding datatypes is:

{
  "name": "reset_defaults",
  "response": {
    "return_code": {
      "message": "string",
      "value": "int16"
    }
  }
}

set_region_of_interest (deprecated)

Persistently stores a 3D region of interest on the rc_visard.

Details

This service can be called as follows.

This service is not available in API version 2. Use set_region_of_interest in rc_roi_db instead.
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/set_region_of_interest

get_regions_of_interest (deprecated)

Returns the configured 3D regions of interest with the requested region_of_interest_ids.

Details

This service can be called as follows.

This service is not available in API version 2. Use get_regions_of_interest in rc_roi_db instead.
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/get_regions_of_interest

delete_regions_of_interest (deprecated)

Deletes the configured 3D regions of interest with the requested region_of_interest_ids.

Details

This service can be called as follows.

This service is not available in API version 2. Use delete_regions_of_interest in rc_roi_db instead.
PUT http://<host>/api/v1/nodes/<rc_itempick|rc_boxpick>/services/delete_regions_of_interest

Return codes

Each service response contains a return_code, which consists of a value plus an optional message. A successful service returns with a return_code value of 0. Negative return_code values indicate that the service failed. Positive return_code values indicate that the service succeeded with additional information. The smaller value is selected in case a service has multiple return_code values, but all messages are appended in the return_code message.

The following table contains a list of common codes:

Table 38 Return codes of the ItemPick and BoxPick services
Code Description
0 Success
-1 An invalid argument was provided
-3 An internal timeout occurred, e.g. during box detection if the given dimension range is too large
-4 Data acquisition took longer than allowed
-8 The template has been deleted during detection.
-10 New element could not be added as the maximum storage capacity of load carriers, regions of interest or template has been exceeded
-11 Sensor not connected, not supported or not ready
-200 Fatal internal error
-301 More than one item model of type UNKNOWN provided to the compute_grasps service
10 The maximum storage capacity of load carriers, regions of interest or templates has been reached
11 An existent persistent model was overwritten by the call to set_load_carrier or set_region_of_interest
100 The requested load carriers were not detected in the scene
101 No valid surfaces or grasps were found in the scene
102 The detected load carrier is empty
103 All computed grasps are in collision with the load carrier
112 Rejected detections of one or more clusters, because min_cluster_coverage was not reached.
300 A valid robot_pose was provided as argument but it is not required
999 Additional hints for application development

BoxPick Template API

BoxPick templates are only available with the +Match extension of BoxPick. For template upload, download, listing and removal, special REST-API endpoints are provided. Templates can also be uploaded, downloaded and removed via the Web GUI. The templates include the dimensions, the views and their poses, if set. Up to 50 templates can be stored persistently on the rc_visard.

GET /templates/rc_boxpick

Get list of all rc_boxpick templates.

Template request

GET /api/v2/templates/rc_boxpick HTTP/1.1

Template response

HTTP/1.1 200 OK
Content-Type: application/json

[
  {
    "id": "string"
  }
]
Response Headers:
 
Status Codes:
Referenced Data Models:
 
GET /templates/rc_boxpick/{id}

Get a rc_boxpick template. If the requested content-type is application/octet-stream, the template is returned as file.

Template request

GET /api/v2/templates/rc_boxpick/<id> HTTP/1.1

Template response

HTTP/1.1 200 OK
Content-Type: application/json

{
  "id": "string"
}
Parameters:
  • id (string) – id of the template (required)
Response Headers:
 
  • Content-Type – application/json application/ubjson application/octet-stream
Status Codes:
  • 200 OK – successful operation (returns Template)
  • 404 Not Found – node or template not found
Referenced Data Models:
 
PUT /templates/rc_boxpick/{id}

Create or update a rc_boxpick template.

Template request

PUT /api/v2/templates/rc_boxpick/<id> HTTP/1.1
Accept: multipart/form-data application/json

Template response

HTTP/1.1 200 OK
Content-Type: application/json

{
  "id": "string"
}
Parameters:
  • id (string) – id of the template (required)
Form Parameters:
 
  • file – template file (required)
Request Headers:
 
  • Accept – multipart/form-data application/json
Response Headers:
 
Status Codes:
Referenced Data Models:
 
DELETE /templates/rc_boxpick/{id}

Remove a rc_boxpick template.

Template request

DELETE /api/v2/templates/rc_boxpick/<id> HTTP/1.1
Accept: application/json application/ubjson
Parameters:
  • id (string) – id of the template (required)
Request Headers:
 
  • Accept – application/json application/ubjson
Response Headers:
 
Status Codes:
  • 200 OK – successful operation
  • 403 Forbidden – forbidden, e.g. because there is no valid license for this module.
  • 404 Not Found – node or template not found