Color Matching Based Approach for Robotic Grasping


Mohammed, Marwan Qaid and Kwek, Lee Chung and Chua, Shing Chyi and Alandoli, Esmail Ali (2021) Color Matching Based Approach for Robotic Grasping. In: 2021 International Congress of Advanced Technology and Engineering (ICOTEN), 4-5 July 2021, Taiz, Yemen.

[img] Text
Color Matching Based Approach for Robotic Grasping.pdf
Restricted to Repository staff only

Download (1MB)


Object grasping is a basic but difficult task in robotic manipulation due to various object size, shapes and other properties. The problem becomes more challenging when the robot has to reach for and grasp a particular object in cluttered environments. To this end, the color of object can be one of the important features for identifying the target object to be picked up. In this paper, we investigate an approach of training a robot so that it is able to locate the target object by matching its color, and goes on to pick up the object in an unsupervised learning manner. The proposed approach is divided into two parts: 1) a semantic segmentation module that locates the target object by segmenting the color image and creates a mask on the predicted target object, 2) an object pose estimation module that predicts the optimal grasp position of target object based on a deep reinforcement learning framework. The proposed approach was evaluated with various testing scenarios of handling single and multiple colored target objects. The experimental simulation results indicate that the suggested approach achieves an overall success rate of 92% in the grasping task.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Image segmentation
Subjects: T Technology > TA Engineering (General). Civil engineering (General) > TA1501-1820 Applied optics. Photonics
Divisions: Faculty of Engineering and Technology (FET)
Depositing User: Ms Nurul Iqtiani Ahmad
Date Deposited: 30 Aug 2021 06:50
Last Modified: 30 Aug 2021 06:50


Downloads per month over past year

View ItemEdit (login required)