Grain bags detection based on improved maximum between-cluster variance algorithm

The key to extracting the edge characteristics of grain bags in the grain reserve warehouse was image segmentation. In practice, the quality of the selected segmentation algorithm directly determined the effect of the bags image segmentation. According to the characteristic of the grain warehouse scene an improved segmentation algorithm based on combining maximum between-cluster variance with edge detection method was proposed to achieve bags edge accurately detecting results as far as possible. First of all, the improved Otsu algorithm which could effectively confirm the bags objective was used to segment the actual scene image initially. And then, comparing with the results of the classical edge detection operator, the bags´ outline could be efficiency extracted by Canny operator. Experimental results showed that the proposed algorithm could effectively extract the bags´ outline, and had the merits of high precision and strong robustness. This work provided a grain bags detection method that laid a good foundation for the further work of bags intelligent identification reckoning.