Rapid automatic capture and tracking detection method of single-frame dark and dim point targets

Abstract

The invention relates to a rapid automatic capture and tracking detection method of single-frame dark and dim point targets, which belongs to the technical field of target rapid capture and tracking in image processing. A point filter is adopted by the invention to detect dark and dim point targets in an image. The detection method comprises the following concrete tracking detection steps of: calling a real-time image and storing an image gray value; designing a point filter to filter the image; sequencing the point targets by utilizing brightness and distance information; recording candidate point target information and carrying out secondary capture; and determining the positions of the point targets and tracking in a small window. The invention has the advantages that the point filter is designed to filter the image, avoiding the problem existing in extraction based on gray level or multi-frame extraction, and realizing target enhancement and background weakening under a complicated sky background, thereby achieving the effect of separating the targets and the background. The targets can be stably detected, and a real-time processing requirement is satisfied, thereby solving the problems of the capture, tracking and measurement of dark, dim and small points. The method has convenient and easy operation, strong real-time property and good stability.

Description

technical field [0001] The invention belongs to the technical field of fast capturing and tracking of targets in images. Background technique [0002] With the advent of the information age, the level of weapons and equipment continues to improve, information countermeasure technology continues to develop, and the requirements for photoelectric tracking and measurement systems are also getting higher and higher. Sufficient response time requires that the target can be detected at a long distance, which is the small target detection problem that has attracted much attention. The performance of small target detection directly determines the effective distance of the photoelectric detection system and the complexity of the equipment, and it plays a pivotal role in the photoelectric system. Using the traditional weak target extraction method, the extraction real-time performance and accuracy are very low, the reasons are as follows: [0003] (1) The traditional point target ex...

AI Technical Summary

Problems solved by technology

[0003] (1) The traditional point target extraction method generally adopts the method of multi-frame accumulation to suppress noise and realize weak target extraction. After the point target appears, real-time capture cannot be performed, and the real-time performance is not high;

[0004] (2) When using threshold segmentation based on brightness, if the background is bright or uneven, it will affect the extraction and tracking of point targets;

[0005] (3) When a large target appears in the background, it will inhibit the extraction of small targets and cause miscapture

[0008] Method ① is the most classic point target extraction method. The advantage of this method is that it is not sensitive to the noise in the image. The disadvantage is that it requires relatively high data storage capacity, and has certain requirements for hardware storage capacity, and it needs previous information for support. , is not conducive to real-time application, and only for the slow moving of the target, it needs to be matched and accumulated between two frames, and it is not suitable for the image with the tracking frame rotating and the background changing rapidly; the method ② generally first performs grayscale statistics on the image, The difference between the target and the background gray level is used to select the threshold value to segment the brighter point target. The advantage of this method is that the amount of calculation is small and the requirements for the hardware are low. The disadvantage is that the performance of the algorithm is affected by the change of the background. Relatively sensitive, not suitable for uneven background or brighter target interference

Images