Nondestructive testing method for solid rocket engine based on computer vision

The invention provides a nondestructive testing method for a solid rocket engine based on computer vision. The method comprises the steps: processing edge pixel points of an image into a series of discontinuous line segments through a data dichotomy according to the gray values, and selecting the vertexes of a polygon or the edge points of an object with large curvature changes in the image as feature points; determining a starting point and an ending point of the defect by utilizing the feature points and a series of discontinuous line segments; and calculating the pixel distance between thestarting point and the ending point, and calculating the actual size according to a certain scale. According to the invention, the problems of low manual detection precision and large human resource investment are solved, manual measurement in the prior art can be replaced, and the measurement efficiency and the measurement precision are greatly improved. 本发明提供一种基于计算机视觉的固体火箭发动机无损检测方法，利用数据二分法将图像边缘像素点根据灰度值大小处理成一系列非连续线段，选择图像