Abstract: In order to do energy saving and emission reduction, this paper designs an adaptive speed control system of coal mine transportation belt based on image recognition. The core of the speed control system is to obtain the coal flow accurately and quickly, so an efficient and accurate coal quantity detection method is designed based on common cameras. The cross section of coal flow is obtained through the registration of reference image and test image, and the area of coal flow area is calculated by Monte Carlo method. The experimental results show that the coal quantity detection is fast and accurate, meeting the real-time requirements of the speed regulation system. The system has the characteristics of low cost of transformation and upgrading, easy maintenance, strong environmental adaptability, and it has broad application prospects in coal mining field.