Research On Binocular Vision Technology For Underwater Robot Operation

With the increasing exploitation of marine resources,it is urgent to improve the underwater operation capability of underwater robot.The binocular vision technology is introduced to meet the needs of ranging and dimensional measurement of underwater targets in underwater operations.Compared with traditional sensors such as radar and infrared,which are expensive and have strict deployment conditions,binocular vision technology can not only accurately measure the target,but also save costs.Therefore,this paper constructs an underwater binocular vision system based on underwater robot,and the main research work is as follows:(1)In the aspect of underwater image processing,the improved CLAHE algorithm is used in this paper to process the obtained underwater image.Firstly,on the basis of in-depth comparison of several typical underwater image enhancement methods,aiming at the problems of excessive brightness enhancement,detail loss and noise enhancement in the normal light region after underwater image enhancement,this paper uses the improved CLAHE algorithm to process the obtained underwater image.Firstly,the color input space is converted from RGB to Lab,and the CLAHE processing is carried out in Lab space.Finally,the color input space is fused back to RGB space for output.Through the comparison of underwater environment image processing effect,it can be clearly seen that after the improved CLAHE algorithm processing,underwater environment image is clearer,more vivid color,higher brightness,contrast,visual effect is more natural,more in line with the real situation of underwater environment.(2)A target ranging and dimensional measurement system based on binocular vision is designed for target ranging and dimensional measurement.Firstly,Zhang Zhengyou calibration method was used to calibrate the camera and obtain the camera parameters.Bouguet algorithm was used for stereo correction of the image.Three commonly used stereo matching algorithms are introduced and analyzed,and SGBM stereo matching algorithm is selected for stereo matching according to the experimental analysis,so as to obtain the depth map.Finally,binocular ranging formula and two-point coordinate formula are used to calculate the distance and size of the target,and obtain the distance and size information of the object in the real scene.Experimental data in the air show that the binocular vision based target ranging and size measurement method has higher accuracy.(3)Carry out the research on binocular vision underwater technology,and establish the underwater target ranging and size measurement system based on binocular vision.During the imaging of underwater camera,the absorption and refraction of water on light will cause the color distortion and contrast reduction of underwater image,which will lead to the inaccurate three-dimensional information obtained by binocular vision.On the basis of underwater image processing,through the analysis of the established underwater camera imaging model,it is found that refraction has an impact on the camera focal length and imaging distortion,and according to the derivation results,the high degree polynomial is used to calibrate and correct the underwater binocular camera distortion.On this basis,SGBM stereo matching algorithm is used to obtain the depth map of underwater scene.Finally,the target distance and size are calculated according to binocular ranging formula and two-point coordinate formula,and the distance and size information of objects in the underwater real scene is obtained.The experimental results show that the system can accurately calculate the real distance and size information of underwater targets in the underwater environment,and the error is less than 6%.