Prediction of successful grasps by end of arm tooling
Given an image and an aligned depth map of an object, the invention predicts the 3D location, 3D orientation and opening width or area of contact for an end of arm tooling (EOAT) without requiring a physical model.
1. A method for determining a successful grasping configuration of an object by an end of arm tooling, comprising the steps of:
receiving from one or more sensors an image of an object, wherein the image includes an aligned depth map;
defining two or more two-dimensional regions on the image, wherein each region includes a plurality of pixels;
dividing each region into two or more sub-regions;
computing features for each pixel within each sub-region to obtain a weight for each sub-region;
adding the weight for each sub-region to obtain a score for each region of the two or more two-dimensional regions on the image;
ranking each region based on the score for each region to obtain a rank for each region; and
selecting the region with the highest rank, wherein the region with the highest rank conveys a successful grasping configuration of the object by an end of arm tooling.
2. The method of claim 1 , wherein the aligned depth map is a disparity image.
3. The method of claim 1 , wherein the aligned depth map is a point cloud.
4. The method of claim 1 , wherein the two-dimensional region is in the form of a rectangle shape.
5. The method of claim 1 , wherein the features for each pixel comprises one or more selected from the group of: color information, texture information, and edge information.
6. The method of claim 1 , wherein the features for each pixel are captured by one or more filters.
7. The method of claim 1 , wherein each region is divided equally into two or more sub-regions.
8. The method of claim 1 , wherein the receiving step further comprises the step of normalizing the image to reduce illumination changes caused by different lighting conditions.
9. The method of claim 1 , wherein the ranking step is performed by a Support Vector Machine (SVM) ranking algorithm.
10. The method of claim 1 , wherein the score is a real number.
11. The method of claim 1 , wherein the end of arm tooling is a jamming gripper.