A Sliding Window Fast Object Detection Method Based on Superpixel Segmentation

Abstract

The present invention relates to a sliding window fast target detection method based on superpixel segmentation, which mainly consists of five parts: image superpixel segmentation, sliding window initialization, sliding window superpixel activation, activation region evaluation and sliding window search based on superpixel activation region constitute. Beneficial effects: the size of the superpixel and the consistency of internal features can be easily changed, thereby changing the superpixel area activated by the same sliding window to achieve the purpose of improving the accuracy of target detection; the superpixel area corresponding to the covered pixel can be included Sliding window, because the feature consistency of superpixels ensures a good segmentation effect on the target boundary, so the present invention can ensure that the activation area and the target boundary have a good and ideal coincidence. The sliding step of the window can be increased within a certain range, while ensuring that the corresponding activation area remains unchanged, so as to achieve the purpose of improving the operating efficiency of the algorithm.

Description

technical field [0001] The invention belongs to the field of image processing and computer vision, and relates to a sliding window rapid target detection method based on superpixel segmentation. Background technique [0002] Vision is an important way for human beings to obtain external information, and images are an important carrier of information. With the development of image processing technology, the size and resolution of images are gradually increasing, and the information contained in them is also constantly enriched. Studies have shown that when humans observe an image, their eyes usually move between the objects contained in the image, and they are not interested in other areas such as the background. Most visual technologies, such as pedestrian detection, face recognition, target tracking and target recognition, etc. , which also acts on the above-mentioned region containing the object. Therefore, how to quickly and effectively detect the object region of inter...

AI Technical Summary

Problems solved by technology

[0006] For the first problem, most of the existing improvements use heuristic algorithms. Firstly, the distribution of the target in the image is preliminarily estimated, and the fine search is performed in the area with high probability, and the rough search is performed in the area with low probability, so as to improve the search efficiency. The purpose of efficiency, but its search accuracy depends on the initial position distribution accuracy, and it will increase the probability of target false detection and missed detection

Although the algorithm based on sliding window search is less efficient, it still has the advantages of high precision and low missed detection rate.

Images