Python - Opencv application example CT image detection edge and internal defects
Applying traditional image processing algorithms to edge detection and defect detection of CT images, the desired effect is as follows:
Regarding image processing algorithms, it mainly involves: grayscale, thresholding, feature extraction such as edges or corners, grayscale similarity transformation, mainly biased towards some 2D geometric transformations, some statistical algorithms involving image matrices, and space curve fitting. Combined, one-dimensional and two-dimensional convolution related content, but to apply them to the actual project to solve the problem, there is a need for another way of thinking in it, analyze from the perspective of the actual image, and deal with the difference in the order of the algorithm It will directly lead to the gap in the final effect, and another unavoidable optimization problem: 调参- -> If the scene is more complex and the collected images are diverse, parameter adjustment is inevitable!
The application example in this paper realizes the detection of the edge and internal defects of CT images. The main algorithms and processes involved are as follows: 灰度->二值化->形态学->感兴趣区域定位->自适应阈值化->剔除异常边缘->绘制结果, and the achieved effects are as follows:
The following describes the overall process of the following algorithm:
1. Introduction to Opencv API
If you want to know the function usage in detail, just search the function name directly, and compare the effect difference of the given algorithm here.
