Mechatronics applications to fish sorting Part 1: Fish size identification

This paper describes research conducted on the design of a new automatic fish sorting system at Monash University, Australia. A concurrent mechatronics approach was implement to achieve the project's goals. Also, a technique based on digital imaging was developed for the task of identification of fish sizes. The system is being implemented in the local fish export industry, providing a low cost operation with improved productivity and reliability in the fish sorting process. A computer-based vision system is used for analysis of the fish. A new algorithm is developed to directly determine the length of fish with fast on-line real-time processing. The measurement technique is also applicable to various species. The design specifications of the vision system and the camera capture devices are also presented in this paper.