Robotic 3-D Laser-Guided Approach for Efficient Cutting of Porcine Belly

Abstract

Using autonomous meat cutting robots to cut porcine belly is a productive and more hygienic solution over traditional, labor-intensive meat cutting practices such as the transmission of the bacteria between the carcasses and manual laborers. Belly cutting is the first step of porcine meat segmentation. The robot needs to be precisely guided to cut open the porcine belly without damaging the guts inside, which facilitates subsequent processing of the guts. However, the body sizes of the pigs are different. As such, the 2-D and 3-D vision methods cannot be employed to identify the belly cutting path. In this article, a laser-guided efficient cutting path planning approach that combines the segmentation algorithm (SA) with an improved genetic algorithm (GA) is proposed. The robotic cutting system consists of an industrial robotic manipulator, customized tools, a linear laser sensor, and a PC. The salient features of the proposed scheme are as follows: 1) Spatial relationships of all the relevant units are established in Cartesian space. 2) Using the eigenvalue decomposition of the covariance matrix, the porcine belly cutting path is recognized from the 3-D laser point cloud of the carcass. 3) Aiming at efficiently optimizing the discrete path waypoints, an adaptive elite GA (AEGA) is derived by improving three operators of GA. 4) In order to alleviate the robotic cutting errors, the SA is designed to shorten the cutting path with no damage to the guts. 5) Under the limit of the peritoneal cavity, the efficient robotic cutting path is optimized by combining the SA with AEGA. Porcine belly cutting experiments are carried out in a porcine automation line under real-world conditions. Experimental results show that the proposed approach is accurate, efficient, and incurs lower costs in the long-term use of the robot.