At present, the particle size analysis of fracturing proppants mainly uses sieving method, while the particle size analysis mainly uses electron microscopy method. The above conventional methods have problems such as long sample delivery distance, low detection efficiency, and large random errors. To this end, research on automatic timed sampling of fracturing site proppants and machine vision detection of fracturing proppant particle size and shape related technologies was carried out. Firstly, a robotic arm automatic timed sampling and machine vision detection system was built to achieve fracturing proppant sampling detection; Secondly, a proppant image recognition algorithm was developed to achieve automatic analysis and calculation of proppant particle size and shape; Finally, experimental comparative analysis was conducted with conventional detection methods. The test analysis results show that compared with the particle size detection using the screening method, the grading curves of the two methods are basically consistent, and the test results are highly consistent; Compared with electron microscopy particle shape detection, machine vision testing has a larger number of particles and no subjective factors, resulting in more accurate and reliable test results; More importantly, compared with conventional detection methods, machine vision achieves on-site detection, significantly improving the number of sample detections and detection efficiency. Research has shown that machine vision methods can solve problems such as long sample inspection distances, low detection efficiency, and large random errors, comprehensively improving the quality control level of fracturing well materials.