Abstract: The synthetic aperture radar (SAR) image of moving target may exhibit phenomena such as shifting and blurring, thereby affecting target positioning and recognition. By estimating the velocity, the phase error of data can be compensated and the moving target can be refocused. However, when the local and overall motion states of the moving target are different, the traditional focusing algorithm exhibits defocusing phenomenon in local micro-motion areas, failing to meet the requirements of high-resolution imaging. Therefore, a sub-region self-focusing method based on target detection is proposed, which estimates the phase error of each sub-region to ensure the overall effect of focusing target. Firstly, a traditional method of refocusing moving target is used for rough focusing; next, cell averaging constant false alarm rate detection algorithm and density-based spatial clustering of applications with noise are used to classify sub-regions, and then local refocusing is performed on sub-regions; finally, the complete target focusing image is obtained by aligning and splicing sub-regions. The effectiveness of the proposed method is verified by processing the real data of spaceborne SAR on aircraft targets.