Application of 3-D Sensor Tracking Imaging in Detailed Feature Extraction of Motion Damage Action

In order to study the theoretical significance and practical value of the application of three-dimensional sensor tracking imaging in motion damage action detail feature extraction, the key technologies such as accurate segmentation of moving targets, regional feature extraction, target description and robust tracking are studied and discussed. The results show that the robustness of particle filter tracking algorithm in complex scenes can be effectively improved by using the separable feature information of the current tracking scene to track the target. From this it can be seen that the Bandelet transform based on geometric flow divides the image into regions, counts the histogram of the intensity distribution of the strip wave in all directions, and extracts the statistical features. Meanwhile, histogram features are normalized in order to maintain the robustness and selectivity of features.