CN-121995339-A - Method, device, equipment and storage medium for detecting moving object in air

CN121995339ACN 121995339 ACN121995339 ACN 121995339ACN-121995339-A

Abstract

The application relates to an air moving target detection method, an air moving target detection device and a storage medium. The method comprises the steps of obtaining single-view complex data, preprocessing the single-view complex data to obtain preprocessed data, conducting sub-aperture segmentation on the preprocessed data to obtain sub-aperture segmentation data, conducting low-rank decomposition and constant false alarm rate detection on the multi-view same-angle sub-aperture data to obtain multi-time phase detection data, conducting low-rank decomposition and constant false alarm rate detection on the single-view different-angle sub-aperture data to obtain single-time phase detection data, and fusing the multi-time phase detection data and the single-time phase detection data to obtain air moving target detection data. The application considers the aerial moving targets which continuously appear at a plurality of imaging moments, and also considers the targets which only appear at a few moments or have sparse time sequence sampling, thereby achieving the effect of improving the extraction capability of the aerial moving targets.

Inventors

SHEN WENJIE
JIA YUNZHEN
WANG YANPING
LIN BIN
LI YANG
BAI ZECHAO

Assignees

北方工业大学

Dates

Publication Date: 20260508
Application Date: 20260130

Claims (10)

1. An airborne moving object detection method, characterized in that the method comprises: acquiring single-view complex data, and preprocessing the single-view complex data to obtain preprocessed data; Sub-aperture segmentation is carried out on the pretreatment data to obtain sub-aperture segmentation data, wherein the sub-aperture segmentation data comprises multi-view sub-aperture data with the same angle and single-view sub-aperture data with different angles; Performing low-rank decomposition and constant false alarm rate detection on the multi-view same-angle sub-aperture data to obtain multi-time phase detection data; Carrying out low-rank decomposition and constant false alarm rate detection on the sub-aperture data of the single scene at different angles to obtain single-phase detection data; And fusing the multi-time phase detection data and the single-time phase detection data to obtain air moving target detection data.
2. The method of claim 1, wherein preprocessing the single view complex data to obtain preprocessed data comprises: Performing logarithmic conversion on the single-view complex data to obtain logarithmic conversion data; carrying out mean value filtering on the logarithmic conversion data to obtain mean value filtering data; performing radiation correction on the mean value filtering data to obtain radiation correction data; and carrying out image registration on the radiation correction data to obtain preprocessing data.
3. The method of claim 2, wherein sub-aperture segmentation is performed on the pre-processed data to obtain sub-aperture segmentation data, comprising: performing Fourier transform on the preprocessed data to obtain Doppler data; performing frequency band division on the Doppler data and constructing a frequency domain weighting function; And carrying out frequency domain weighting and inverse Fourier transformation on the Doppler data according to the frequency domain weighting function to obtain sub-aperture segmentation data.
4. The method of claim 1, wherein the performing low-rank decomposition and constant false alarm rate detection on the multi-view same-angle sub-aperture data to obtain multi-temporal phase detection data comprises: Expanding the multi-view same-angle sub-aperture data into column vectors according to pixel positions, and sequencing according to time sequence to obtain multi-view same-angle sub-aperture matrix data; Performing low-rank decomposition on the multi-view same-angle sub-aperture matrix data to obtain multi-view sparse sequence data; And performing constant false alarm rate detection on the multi-scene sparse sequence data to obtain multi-time phase detection data.
5. The method of claim 4, wherein the performing low-rank decomposition on the multi-view identical-angle sub-aperture matrix data to obtain multi-view sparse sequence data comprises: decomposing the multi-view same-angle sub-aperture matrix data into a low-rank part and a sparse part according to a low-rank sparse decomposition model; Solving the low-rank part and the sparse part according to a preset low-rank constraint parameter and a preset sparsity constraint parameter to obtain multi-view sparse item data; Rearranging the multi-scene sparse item data according to columns to obtain multi-scene sparse sequence data.
6. The method of claim 5, wherein the performing constant false alarm rate detection on the sparse sequence data to obtain multi-temporal phase detection data comprises: Setting pixel data to be detected in the sparse sequence data as a target area, and determining a background area according to the target area; Calculating a probability density function of the background area, and obtaining a detection threshold value according to a preset false alarm probability value and the probability density function; obtaining pixel detection data according to the pixel data to be detected and the detection threshold value; And merging the pixel detection data according to the space position to obtain multi-temporal detection data.
7. The method of claim 6, wherein the performing low-rank decomposition and constant false alarm rate detection on the single view different angle sub-aperture data to obtain single phase detection data comprises: expanding the sub-aperture data of the single scene at different angles into column vectors according to pixel positions, and sequencing according to the sub-aperture index sequence to obtain sub-aperture matrix data of the single scene at different angles; performing low-rank decomposition on the single-view sub-aperture matrix data at different angles to obtain single-view sparse sequence data; and carrying out constant false alarm rate detection on the single-scene sparse sequence data to obtain single-phase detection data.
8. An airborne moving object detection device, said device comprising: the preprocessing unit is used for acquiring single-view complex data, preprocessing the single-view complex data and obtaining preprocessed data; the sub-aperture segmentation unit is used for carrying out sub-aperture segmentation on the preprocessing data to obtain sub-aperture segmentation data, wherein the sub-aperture segmentation data comprises multi-view sub-aperture data with the same angle and single-view sub-aperture data with different angles; The multi-temporal detection unit is used for carrying out low-rank decomposition and constant false alarm rate detection on the multi-scene same-angle sub-aperture data to obtain multi-temporal detection data; the single-phase detection unit is used for carrying out low-rank decomposition and constant false alarm rate detection on the sub-aperture data of different angles of the single scene to obtain single-phase detection data; And the fusion result unit is used for fusing the multi-time phase detection data and the single-time phase detection data to obtain air moving target detection data.
9. A computer device, comprising: At least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores computer instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.
10. A computer readable storage medium having stored thereon computer instructions for causing a computer to perform the method of any of claims 1 to 7.

Description

Method, device, equipment and storage medium for detecting moving object in air Technical Field The present application relates to the field of object detection technologies, and in particular, to a method, an apparatus, a device, and a storage medium for detecting an airborne moving object. Background Synthetic Aperture Radar (SAR) is well known as an active microwave imaging radar system. The device is arranged on a motion platform (such as a satellite, an airplane and an unmanned aerial vehicle), and an equivalent ultra-long antenna is synthesized by transmitting electromagnetic wave pulses and receiving echo signals reflected by ground objects by utilizing an advanced signal processing technology, so that a high-resolution two-dimensional image (distance direction and azimuth direction) is obtained. The device has the greatest characteristics of all-day and all-weather working capacity, can penetrate through cloud, rain and snow and ground vegetation to a certain extent, and is an important supplement of optical remote sensing. In the traditional implementation mode, the moving target detection of the single-channel SAR data focuses on ground or sea surface scenes in a multi-way mode, most methods are based on full-aperture imaging result processing, and common means comprise threshold segmentation of single-scene images, multi-time phase amplitude difference, simple change detection and the like. The aforementioned method generally treats each image as an observation of a stationary scene, and finds an abnormal region by comparing amplitude differences or local statistical properties of the images at different times. However, in the above-mentioned manner, the motion characteristics of the aerial moving object are not specially processed in the azimuth imaging process, so that the aerial moving object is easy to generate azimuth defocusing in the full-aperture imaging result, and the signal-to-noise ratio of the object is reduced. On the basis, threshold segmentation or multi-time phase difference time division is carried out, so that the contrast between the aerial moving target and the complex background is insufficient and is easily submerged by the strong scattering background. The method can play a role under the condition that the background is relatively simple or the target change is obvious, but for an air moving target in a satellite-borne SAR image, an ideal detection effect is difficult to obtain only by means of a full-aperture single-view image or a simple multi-time phase difference due to the fact that the target size is small, signals are weak and the motion state is complex. Disclosure of Invention Based on the method, the device, the equipment and the storage medium, the method, the device, the equipment and the storage medium for detecting the moving target in the air are provided, and the moving target in the air is obtained by acquiring single-view complex data, preprocessing the single-view complex data, sub-aperture segmentation, low-rank decomposition, constant false alarm rate detection, single-multi-time phase detection data fusion and other operations. The method not only considers the aerial moving targets continuously appearing at a plurality of imaging moments, but also considers targets which only appear at a few moments or have sparse time sequence sampling, thereby achieving the effect of improving the extraction capability of the aerial moving targets. In a first aspect, there is provided an airborne moving object detection method, the method comprising: acquiring single-view complex data, and preprocessing the single-view complex data to obtain preprocessed data; sub-aperture segmentation is carried out on the pretreatment data to obtain sub-aperture segmentation data, wherein the sub-aperture segmentation data comprises multi-view sub-aperture data with the same angle and single-view sub-aperture data with different angles; Carrying out low-rank decomposition and constant false alarm rate detection on the multi-view same-angle sub-aperture data to obtain multi-time phase detection data; carrying out low-rank decomposition and constant false alarm rate detection on the sub-aperture data of different angles of a single scene to obtain single-phase detection data; And fusing the multi-time phase detection data and the single-time phase detection data to obtain the air moving target detection data. According to one implementation manner in the embodiment of the present application, preprocessing is performed on single-view complex data to obtain preprocessed data, including: carrying out logarithmic conversion on the single-view complex data to obtain logarithmic conversion data; average filtering is carried out on the logarithmic conversion data to obtain average filtering data; performing radiation correction on the mean value filtering data to obtain radiation correction data; And carrying out image registration on the radiation correction data to obtain preprocessin