On-line Measurement Of Aircraft Skin Hole Size Based On Binocular Vision

Machining precision detection of aircraft skin hole is very important for automatic assembly of workpiece in aviation field.The aircraft skin is a thin-walled structure with relatively small edge characteristics and high detection accuracy index.The manual line cutting and film sticking method is prone to misdetection due to its low detection efficiency and high intensity operating environment,and at the same time,there are some safety risks.At present,the level of automatic detection of skin holes in China’s aircraft manufacturers is generally low,which is more unable to meet the actual needs.The extraction of geometric parameters of skin holes in workpieces and the identification and detection of position and pose still need to be explored and practiced continuously.Aiming at the above problems,this paper designs a set of binocular stereo vision detection scheme for real-time and accurate measurement of aircraft skin hole size.Based on the design of the visual module of the end effector of the transplanting robot automatic drilling and riveting detection system,the overall technical route is given with the skin hole in the single curved skin as the research object.The process covers the implementation of the hardware and software system.The main content is divided into the following four points:Firstly,according to the measurement precision index and measurement method of the skin hole,the overall scheme was given,the camera linear and distortion models were established and analyzed,and the coordinate transformation of the center point of the skin hole from the pixel coordinate system to the world coordinate system was realized.Secondly,establish the binocular system structure model based on triangle range,by mathematical model method to derive the skin hole and three-dimensional coordinate of the center for structural parameters of the involved,and calculate the error transfer coefficient analysis meet the requirement of accuracy in the measurement system parameters inside and outside value range,as a standard of measurement system hardware selection,Through the reasonable construction of hardware platform to complete the design of hardware detection system;Thirdly,the software system of skin hole measurement is designed.The gray histogram was optimized according to the image features,and a bilateral filter was used to denoise the image.In feature detection,Canny edge detection operator was optimized to extract edge feature points by adaptive threshold.The fitting circle algorithm was applied to the binocular system to obtain the coordinates of the hole center and the aperture with sub-pixel accuracy.Based on this,the software interface of the skin hole image processing system was prepared.At the same time,the sequential detection method of the group holes in the mask was designed.Finally,the internal and external parameters of the camera were obtained by high-precision binocular calibration experiments based on the Halcon hexagon array calibration plate,and the left and right viewing angles were corrected by polar constraint to perform image matching and depth calculation.Two calibration and simulation experiments were carried out by adjusting the test bed to realize online positioning and 3D measurement of skin holes.