Objective To address the prominent issues of delayed feedback and low detection efficiency in the offline particle size distribution detection of traditional wheat processing.Methods The dynamic image method was applied to particle size detection in the wheat processing process. An online particle size detection device was designed based on wheat processing technology, including the design of key structures and algorithm software. The device was compared and analyzed with the manual sieving method for sample materials.Results Compared with the sieving method and laser method, the maximum relative errors in the detection results of the device for the I and II skin grinding materials were no more than 3.2%, 14.7%, and 2.3%, 14.2%, respectively. The relative error of the repeatability test for the I skin grinding material did not exceed 2.6%.Conclusion The application of the dynamic image method to online particle size detection in the wheat processing industry is feasible.