Video target detecting and tracking method based on optical flow features

Abstract

The invention provides a video target detecting and tracking method based on optical flow features. According to the technical scheme of the method, during the first step, an input image frame sequence is subjected to background sampling, and the optical flow vector of each pixel point after the sampling process is calculated. Meanwhile, the background motion is estimated based on the Mean Sift algorithm, and then the overall significance of a target is estimated. Finally, a threshold value is set according to the detection result of the target significance detection, so that a target region and a background region are separated. During the second step, the tracking of a video target is conducted: firstly, the target region is selected as a positive sample, and the background region is selected as a negative sample. The target is described based on the Haar features and the global color features of the target. Meanwhile, original features are subjected to sampling and compressing in the random matrix manner. Based on the bayesian criterion, the similarity between the target and a target of a previous frame is judged. Finally, the target is continuously tracked based on the particle filtering algorithm. In this way, multiple features including the target motion saliency, the color, the texture and the like are fused together, so that the success rate of target detection is improved. Therefore, the target can be quickly, effectively and continuously tracked.

Description

technical field [0001] The invention belongs to the technical field of visual image processing, and relates to a target detection and tracking method, in particular to a video target detection and tracking method based on optical flow features. Background technique [0002] The detection and tracking of moving objects is one of the hotspots and difficulties in computer vision research. The detection and tracking of moving objects has broad application prospects in intelligent transportation, security monitoring and military fields. UAVs have the advantages of high maneuverability, high resolution, good concealment, and flexible operation. Using the video sensor carried by UAVs to track and analyze ground moving targets has important practical significance and theoretical value. [0003] However, due to the movement of the video sensor with the high-speed movement of the UAV, the complexity of the background in the video sequence images and the diversity of moving target inf...

AI Technical Summary

Problems solved by technology

The feature-based tracking method achieves target tracking by extracting the salient features of the moving target. This algorithm is not affected by the deformation of the target or the change of illumination, and it can also achieve tracking when the target is partially occluded. The continuous tracking of the moving target is difficult; based on the area tracking method, the moving target detection method is first used to segment and extract the region containing the moving target, and a template is formed to represent the region with a rectangle or an ellipse, and then the grayscale or color-related algorithm is used for the next step. The frame is matched to achieve tracking. This algorithm realizes stable tracking under the premise that the target is not occluded. When the target deforms or the occlusion area is large, the tracking accuracy decreases; the contour-based tracking method is to track the target boundary contour, using The global information of the target contour obtains a closed curve without prior knowledge of the target, but this method requires manual intervention for initialization, and the calculation is very time-consuming; based on the model tracking method, the target shape model and motion model are first established, and then according to the actual video Sequence images determine model parameters and motion parameters to achieve target tracking, but it is very difficult to obtain accurate real models of targets in actual scenes

Images