Key of packaged granary grain quantity recognition —Grain bags image processing

摘要

A study of core problem in the packaged granary grain intelligent detection based on image recognition was conducted, and an intelligent detection method combining Fisher criterion with Adaptive Genetic Algorithm was used to solve it. We took the actual scene video as the analysis object, and the constructed Fisher criterion as the fitness function of Genetic Algorithm, while we presented a local optimization operator that enhance the diversity of the population and solve the disadvantages of poor astringency and premature occurrence. Finally, we introduced the morphology processing method to achieve excellent detection effect. Experimental results showed that this detection algorithm effectively improves the anti-jamming capability and robustness. This work provides an intelligent detection method for grain reserving management.