Research on Parallel Target Detection Technology Based on Sea Land Segmentation Results

This combined with the data processing requirements of large-scale sea land mixed high-resolution remote sensing images, this paper proposes a parallel target detection algorithm based on sea land segmentation results, and uses mpi4py function library to realize the parallel algorithm. The large-scale high-resolution remote sensing images are divided into smaller sub image blocks, and multi process synchronous sea land segmentation processing is realized based on the trained semantic segmentation model, In addition to the whole ocean sub image block, only the whole land sub image block and sea land mixed sub image block are subject to target detection processing, so as to reduce the target detection area of high-resolution remote sensing image, greatly reduce the data processing time of large-scale highresolution remote sensing image, and improve the detection efficiency. The experimental results show that on the basis of ensuring the detection accuracy, the parallel segmentation before detection technology can significantly shorten the image processing time and has better speedup and scalability than the traditional parallel target detection technology.