The commonly used shrimp peeling equipment is roll shaft shrimp peeling machine. However, the machine can not adapt to shrimp material feeding information in real time and lacks automatic control system. Aiming at the problems of poor material adaptability, low degree of automation, lack of parameter acquisition ability existing in the roller shrimp peeling machine, the material information detection method and equipment operation parameter detection method based on machine vision technology were studied, and the parameter detection and control system of the roller shrimp peeling machine was designed. The system was mainly composed of image acquisition module, parameter detection module and control module, which can realize the real-time detection and optimal control of the main parameters of peeling machine. By analyzing the relationship between the number of shrimp pixels and the quality of shrimp, combined with the range of shrimp size, the detection model of shrimp size and feeding rate was proposed. According to the working characteristics of the peeling machine, the detection methods of roll shaft rotation angle and speed, tappet frequency, raw material lifting belt speed and water flow rate were proposed. The results of parameter monitoring test showed that the monitoring errors of shrimp size, feeding rate, roll rotation angle, roll rotation speed, tappet frequency, raw material lifting belt rotation speed and water flow rate were 0, 3.46%, 0.51%, 1.73%, 1.93%, 3.34% and 0.92%, respectively. The control errors of roll rotation angle, roll ratation speed, tappet frequency and raw material lifting belt ratation speed were 0.53%, 1.04%, 2.15% and 3.34%, respectively.