UAV (unmanned aerial vehicle) reconnaissance video image object automatic detection method

Abstract

The invention relates to the field of image processing, and specifically relates to a UAV (unmanned aerial vehicle) reconnaissance video image object automatic detection method. The method comprises the steps: employing a target coarse extraction algorithm and a fine extraction algorithm for pre-extraction of a target region, carrying out the judgment of the pre-extracted region through a classifier, eliminating a non-target pre-extracted region, maintaining and marking the target region, and achieving the automatic detection of a target in the video image. According to the invention, a pre-extraction algorithm is introduced for the pre-extraction of the target in the video image, thereby avoiding a condition that a detection algorithm causes the detection leakage or wrong detection because the size of a sliding window and the step length cannot achieve the stepless changes or the features of the target are not apparent, and improving the detection accuracy. Meanwhile, the invention introduces the pre-extraction algorithm for the pre-extraction of the target in the video image, thereby achieving the sharing of the weight of a classifier detection algorithm in the whole automatic detection process of a reconnaissance target, reducing the requirements for the number of samples required by the training of the classifier and the order of magnitudes of diversity, and improving the operation efficiency.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method for automatically detecting objects in unmanned aerial vehicle reconnaissance video images. Background technique [0002] In recent years, UAVs, as an emerging combat force, have played an irreplaceable role under the conditions of informationized operations. Vigorously developing UAV equipment technology is of great strategic significance for improving the combat capabilities of troops. As one of the key technologies for UAVs to perform reconnaissance and strike missions, automatic target detection technology can provide strong support for technologies such as UAV visual navigation, target tracking, autonomous flight and damage assessment. Efficient and accurate detection algorithms can effectively reduce the burden on ground operators, improve rapid response combat capabilities and system detection capabilities. [0003] At present, among image target detection algorit...

AI Technical Summary

Problems solved by technology

[0005] 1) The traditional target detection algorithm retrieves the target by traversing the entire image through sliding windows, which not only affects the detection efficiency, but also affects the detection accuracy. Based on the existing hardware processing, it is difficult to guarantee real-time performance;

[0006] 2) Manually collecting and marking a large number of samples is time-consuming and laborious, the training time is long, and it is difficult to consider the diversity of samples;

[0007] 3) UAV reconnaissance video images contain a lot of redundant information, the number of pixels occupied by the target is insufficient, and the proportion in the whole picture is small, the texture features are not obvious, and it is easy to cause false detection or missed detection

Images