Application Of Underwater Robot Detection Technology And Visual Identification Of Leakage Points

With the increasing demand for economic and social resources and the improvement of science and technology,marine resources,especially underwater oil and natural gas,are being considered more seriously.In the offshore oil and gas industry,underwater LPG pipelines are indispensable components of offshore oil and gas field mining and transportation,and they need to be tested in time to ensure production safety.In todays world where more and more attention is paid to underwater detection,underwater robots can be used to inspect the integrity of underwater LPG pipelines to prevent leakage accidents and potential safety hazards.Aiming at the purpose of inspection and detection of underwater LPG pipelines,this paper designs a bionic fish-type underwater robot that can collect underwater images and transmit signals with the host computer.The underwater robot is mainly composed of STM32 single-chip microcomputer as the core controller,COMS serial camera VC0706,single-serving machine SG90 and communication module ESP8266,which are equipped with ?C/OS-III real-time operating system.The underwater image acquired by the COMS camera is uploaded by the main control chip STM32 to the local host computer via the ESP8266 module.The host computer saves the source image and copies the image copy for image preprocessing.This study will use the source computer vision library OpenCV for grayscale correction,image enhancement and edge detection in turn.The spatial domain bilateral filtering algorithm is used to maximize the degradation of underwater images and improve image clarity.The edge detection of the grayscale corrected image is performed by the Scharr filter.Finally,a hardware test platform for simulating shallow water environment is built according to the module structure of this paper.The QT framework is used to design a user interface that includes image comparison,signal transmission,and motion steering.The system debugging shows that the underwater robot can wirelessly transmit the underwater oil leakage image to the local PC end,and analyze the image preprocessing effect in combination with theimage sharpness evaluation standard.The image sharpness is greatly improved,the oil bubble edge can be protruded,and the cost performance is improved.High,it proves that the underwater robot designed in this paper is feasible.