Moving Object Recognition on Production Line Based on Adaptive Frame Differencing Algorithm

In order to solve the problem that the traditional moving object recognition algorithms for industrial production lines are susceptible to external interference in complex backgrounds and the recognition accuracy is reduced, this paper proposes a moving object recognition algorithm based on Adaptive Frame Differencing (AFD). Firstly, the contrast of moving object images on the production line is enhanced by using gamma transformation for image enhancement and bilateral filtering for noise reduction. Then, the AFD algorithm is used to combine the global distribution of image pixels to realize the dynamic adjustment of image threshold. Based on adaptive threshold, the proposed algorithm can achieve image segmentation and accurate recognition of moving objects under different backgrounds and external interference conditions. In order to verify the reliability of the algorithm, three sets of moving object image datasets under different conditions are tested. The accuracy and recall rate indicators are significantly improved.