An identification and localization approach of different labware for mobile robot transportation in life science laboratories

Abstract

The accurate detection and position estimation of objects is vital to perform transportation tasks safely with mobile robots. This work presents an approach to identify and localize multiple labware which contains chemical and biological components to be manipulated and transported in life science laboratories using H20 mobile robots. The H20 robot has dual arms where each arm consists of 6 revolute joints with 6-DOF and 2-DOF grippers. The Kinect sensor is used to detect the required labware on a wide workstation to be manipulated wherever it is located. Polarization and intensity filters are used with the Kinect to eliminate the effects of strong and glossy light on the identification process. The local features recognition based on SURF algorithm (Speeded-Up Robust Features) is used to differentiate multiple labware which are defined by some marks. Some preprocessing steps are applied to the image to improve the recognition process. A localization procedure is performed for the target using the Kinect depth frame with a mapping process. The implemented approach approves an efficient performance with a high success rate under different lighting conditions.