Objective: In order to improve the mechanization and automation level of catfish head cutting. Methods: Through the study on the shape parameters and friction characteristics of catfish, the related parameters of catfish were obtained. Based on the machine vision technology to identify the assembly line fish, the end of the fish skull was defined as the fish head cutting track line, and an adaptive adjustment device was designed to adjust the position of the fish body, so that the disc cutter can avoid the skull for head removal. The GUI was designed in order to achieve visual control. Results: The system test results showed that the success rate of cutting track line discrimination of the image acquisition system reached 90%, with the average error of the adaptive adjustment system of 0.21 cm, which confirmed the reliability of the system. Conclusion: According to the preset processing efficiency, about 600 catfish can be processed per hour.