Fishing net real-time autonomous detection method for underwater obstacle avoidance system

Abstract

The invention provides a fishnet real-time autonomous detection method for an underwater obstacle avoidance system, and belongs to the field of computer vision, and through an underwater laser detection means and a method based on combining of a depth generation network and a deep learning regression method with regional suggestions for target detection, the time for image pre-processing operationcan be reduced. An underwater laser scanning system is used to collect a fishnet images, and overcomes the underwater back scattering effect of traditional light and the problem that sound waves aresensitive to seawater environment, clear and high brightness underwater fishnet images can be obtained, and can be directly used in later stage fishnet detection without pre-processing methods such asdenoising, the real-time performance of the later stage fishnet detection process can be greatly improved, the requirement of a deep learning large data volume can be ensured, the confidence of the fishnet detection can be improved while ensuring of the real-time performance, and a fishnet detection technology method for visual part for the underwater obstacle avoidance system is provided.

Description

technical field [0001] The invention provides a real-time autonomous detection method for fishing nets used in an underwater obstacle avoidance system, which belongs to the field of computer vision, specifically through underwater laser detection means and a method based on depth generation network and deep learning regression and area It is recommended to combine the method of target detection to reduce the time of image preprocessing operation, and the dynamic process of fishnet frame regression can still maintain good time performance under the premise of increasing confidence. Background technique [0002] In the past ten years, my country's marine fishing industry and aquaculture industry have achieved unprecedented rapid development, which has played a positive role in increasing foreign trade exports and increasing economic income. However, with the development of marine fishing and aquaculture, more and more incidents of fishing nets being entangled with propellers h...

AI Technical Summary

Problems solved by technology

[0004] The invention provides a real-time autonomous detection method for fishing nets used in an underwater obstacle avoidance system, the purpose of which is to detect the traditional light in the detection technology for underwater targets due to the serious backscattering effect underwater, resulting in a relatively short detection distance , and sound waves are more sensitive to environmental factors to improve these problems

Images