SAR image ship detection method and system based on lightweight deep learning

Abstract

The invention discloses an SAR image ship detection method and system based on lightweight deep learning, wherein the method comprises the steps: carrying out the preprocessing of a large-size SAR image, and selecting a training sample; introducing a Ghost module and Ghost Bottleneck to upgrade the YOLOv5s to obtain a preliminary lightweight model of the YOLOv5s; on the basis of the preliminary lightweight model, utilizing the network pruning and knowledge distillation of a traditional model lightweight algorithm to realize further lightweight of the model; using a TensorRT reasoning optimizer for carrying out reasoning acceleration on the lightweight YOLOv5s model, and deploying the lightweight YOLOv5s model on NVIDIA Jetson TX2; cutting large-size SAR images to be detected, and sequentially sending the cut large-size SAR images to the model to complete detection; and synthesizing a detection result, and screening a prediction frame on the final large-size SAR image by using NMS non-maximum suppression. On the premise of satisfying acceptable precision loss, the parameter quantity and the floating point operand of the model are compressed, and the detection speed is improved.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a SAR image ship detection method and system based on lightweight deep learning. Background technique [0002] Alexnet was born in 2012, setting off a boom in the application of deep convolutional neural networks in the computer field. A deeper model often means that the model has better nonlinear expression ability, can complete more complex transformations, and thus can fit more complex features. Based on such an assumption, the deep convolutional neural network is developing towards deeper and wider directions. Although it shows better performance in various tasks, the volume of the network model follows The amount has become increasingly large, which is contrary to the current hardware conditions of various embedded devices on the mobile terminal. The various achievements of deep neural network research can only be shelved and cannot be implemented. Comp...

AI Technical Summary

Problems solved by technology

[0005] High-resolution image target detection based on big data has always been a hot research direction in the field of remote sensing image processing. Traditional target detection and recognition methods cannot be adaptively adjusted for the massive data of remote sensing images, and a large number of image features need to be artificially designed. At the same time as the huge time cost, it puts forward extremely high requirements for researchers in terms of professional knowledge and understanding of data characteristics, and searching for efficient classifiers to fully understand the data is like finding a needle in a haystack

Images