The Research Of Aircraft Skin Hole Geometry Parameters Detection Technology Based On Machine Vision

The manufacture precision of aircraft skin holes has a great impact on theservice life and connection strength of the aircraft mechanical structure. Innowadays the detection means of the aircraft skin holes is relatively insufficient inChina. In most cases the detection is taken by human eye or simple measuring witha plug gauge to the holes of important parts. No special testing equipment isadopted to measure the specific size of the holes, so the accuracy and efficiencycan’t satisfy the demands. Generally the skin holes are small, usually from5mm to6mm and they are hard to measure with common measuring tools, which hadlimited the development of aircraft assembly technology. Therefore it is urgent toput forward a rapid and accurate measurement system.This paper focuses on the geometry parameters detection of aircraft skin holesin small curvature surface based on machine vision technology. The overallscheme of measurement system is designed according to the manufacture accuracyof the skin holes. Based on the study of the reference of measuring system, the axisof skin hole should be determined during the process of drilling the hole. Oneself-calibration algorithm is proposed for the installation of infrared distancesensor used to find the skin hole axis, which can ensure the installation position iscorrect.The CCD camera and optical lens are selected according to the measurementaccuracy and the size of optical window. Through compare and analysis this paperchooses the CCD camera and optical lens whose resolution is below the theoreticalcalculation the system requirement, and the resolution can be compensated withsub-pixel edge location algorithm. The light source of the measurement system isanalyzed and the ring light is finally chosen in order to eliminate the shadow in theimage. The image acquisition card is selected which can correctly match thecamera model.The preprocessing algorithm and pixel edge detection algorithm of skin holesimage are studied. The grey value equalization of graphics processing is applied tothe skin hole image for enhancing the image contract. The combination ofmorphological edge detection operator and Canny edge detection operator is usedto obtain the pixel level edge of aircraft skin hole image The spatial momentsub-pixel edge subdivision algorithm is applied to the skin hole image to get thesub-pixel edge point coordinates.Skin hole image data processing method is studied including filtering invalidedge points using ROI method, fitting for the edge point and analyzing the fittingerror using least square method and calculating the roundness of the least square fitting circle. The skin hole is measured with the program in this paper and themeasuring results are compared with the standard size of the same skin hole. Theexperimental results reveal the measurement system designed in this paper cangive out comparatively accurate geometry size of the aircraft skin hole. Finally thevisual interface convenient for user operation is complied with MFC and Opencv.