Ship target detection method for SAR image based on image block

• Published in: 2023 9th International Conference on Computer and Communications (ICCC) pp. 2057 - 2061
• Main Authors: Dongjuan, Han, Qiang, Xue, Gang, Li, Aiwen, Liu, Feiyang, Xue, Shuai, Yuan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 08.12.2023
• Subjects: Hardware; image blocks; Image segmentation; multi-core digital signal processor; Object detection; Orbits; Radar polarimetry; real-time detection; Real-time systems; SAR image; ship detection; Signal processing algorithms
• DOI: 10.1109/ICCC59590.2023.10507300

In order to solve the problems of large amount of SAR image, large differences in target resolution, non-uniform sea clutter, and difficulty in real-time detection in orbit, an Algorithm for ship detection in spaceborne SAR image based on image segmentation is proposed. Firstly, the image is processed into blocks according to the azimuth and distance directions, secondly, the target detection method based on Weibull statistical model is used to detect ship targets on each block, and finally the connected domain labeling method is used to realize the target clustering. The algorithm effectively solves the problem of target segmentation, improves the detection efficiency and precision, and reduces the false alarm rate. The test is carried out on the hardware platform of multi-core DSP-TMS320C6678 + XilinxV7XC7VX690T. The test results show that the detection effect and efficiency are improved.