Basler Cameras Are at the Heart of an Automatic Fish Sorting and Identification System
In Frutigen, located in the Swiss Alps, warm water drained by the Lötschberg railway base tunnel is used for the sustainable production of sturgeon. To achieve optimal growth conditions, the fish must be periodically sorted. Also, the sturgeon must be individually identified based on their caviar production. The first goal of the project was to develop a device that automatically sorts sturgeon while they are swimming naturally through an underwater monitoring channel. Compared to conventional manual sorting methods, this saves costs, is stress-free for the fish, and reduces the risk of injuries. The second goal was to investigate the possibility of individually identifying sturgeon based on biometric features. In the future, this could replace the use of RFID chips. An individual identification system requires the existence of a measurable, finger-print-like feature on the sturgeon. The patterns on the sturgeon’s forehead and nose are considered to have the required individual characteristics. Therefore, a second vision system consisting of a Basler piA2400-17gc camera and an LED illumination panel was built up and mounted in a waterproof housing. The high resolution of five megapixels is required for capturing the fine structure within the forehead and nose patterns. The development of appropriate image processing algorithms for individual identification will need additional work. The Gigabit Ethernet connections on both camera models provide high flexibility in placing the image processing computer. The computer is typically located at a 20 meter distance from the water.