Object surface deformation feature extraction method based on line scanning three-dimensional point cloud

A method for extracting surface deformation features of an object based on a linear scanning three-dimensional point cloud, comprising:

• step 1:
• performing data acquisition by using a three-dimensional measurement sensor based on line structured light scanning to realize synchronous measurement of sectional profiles at same attitudes and at same times;
• step 2:
• preprocessing the sectional profiles of the object measured by the three-dimensional measurement sensor by correcting a systematic error due to an installation deviation of the three-dimensional measurement sensor and a radian of laser lines in the measurement through a calibration file;
• step 3:
• extracting sectional main profiles from the preprocessed sectional profiles of the object one by one;
• step 4:
• obtaining features of the type of a large area deformation by analyzing a deviation between the sectional main profiles and standard profiles and obtaining features of the type of a small area deformation by analyzing a deviation between the preprocessed sectional profiles and the sectional main profiles, based on characteristics of the sectional profiles;
• and extracting deformation feature points of the sectional profiles in combination with a deformation feature knowledge base;
• step 5:
• forming a characteristic binary image by the deformation feature points, and making statistics on lengths and geometries of various connected regions in the characteristic binary image in combination with the deformation feature knowledge base, and then dividing the characteristic binary image into image sub-blocks which do not overlap with each other;
• for each image sub-block, labeling the image sub-block as a deformation target morphology sub-block if the image sub-block contains a long connected region or has feature points that have target morphological characteristics;
• step 6:
• performing a morphological operation on deformation feature points in a set of deformation target morphology sub-blocks, removing noise regions of short-length, and generating a deformation region of confidence ROC;
• and then performing a region growth to extract a target using the deformation region of confidence;
• step 7:
• making statistics on deformation characteristic values of a deformation region in the surface of the object, including linear characteristic values, area array characteristic values, and degrees of deformation, according to predefined deformation features.