Abstract
Pose estimation that locates objects in a bin is necessary for a robotic bin picking system. Although many algorithms have shown high performance in pose estimation, most algorithms estimate the poses of objects regardless of their occlusion. This can reduce the success rate in picking up the object. To resolve this issue, we propose a novel pipeline that estimates a pose only for occlusion-free objects based on point pair feature-based pose estimation with multiple edge appearance model (PPF-MEAM). The proposed method detects occlusion-free objects in the 2D image captured by a camera with a convolutional neural network framework. Next, corresponding point clouds of occlusion-free objects need to be extracted by using their locations in the 2D image. we propose a robust extraction method that finds the 3D points corresponding to image pixels in the 2D image to reduce the effect of the calibration errors between the camera and 3D sensor. The point cloud of the occlusion-free objects is finally input into a pipeline of PPF-MEAM to estimate the pose of the object. The experiment results prove that the proposed method is about 50% faster 30% higher in terms of pose estimation success rate compared with the original PPF. Moreover, it increases the success rate of picking tasks compared with the original PPF-MEAM.
Highlights
The last decade has seen a rapid increase in the introduction of industrial robots to production lines to increase their productivity and overcome labor shortages
This method uses 3D point cloud data and 2D image data of the scene as the input; it has shown high performance in pose estimation experiments. After integrating it into a robotic bin picking system, we found that it cannot help the system achieve a high success rate because point pair feature (PPF)-MEAM estimates the poses of the objects regardless of their occlusion, which will likely result in the robot picking an occluded object in the bin
REAL SCENE DATA We evaluated an occlusion-free PPF-MEAM using data of real scenes, the results of which are described
Summary
The last decade has seen a rapid increase in the introduction of industrial robots to production lines to increase their productivity and overcome labor shortages. Some tasks conducted in an unstructured environment are still difficult to automate. A robotic bin-picking task is one such challenging task. To achieve the robotic bin-picking, the robot must locate an object placed among other objects in an unstructured environment, where the objects change their positions and orientations every time an object is removed from the bin. Locating a distinct object in a bin, in other words, pose estimation of a distinct object, has become possible with the introduction of commercialized three-dimensional (3D) sensors. With rapid advances in 3D measurements [1], [2], such as stereo vision [3]–[5], active stereo methods [6], [7], time
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call