Image processing for transmission line extraction and detection

Abstract

This dissertation presents the image processing for transmission line extraction and detection. It includes the procedure to realize the size measurement of the transmission line on PCB to assist the electromagnetic analysis. The project is developed with python and opencv comprising three parts, which are board extraction, transmission line detection and size measurement. The whole process involves the color model conversion (from RGB to HSV, RGB to grayscale), image smoothing (bilateral filtering), thresholding (OTSU adaptive threshold), image enhancement (morphological operation), contour findings, edge detection (Canny edge) and segmentation measurement. Through a combination of those methods in each step, we finally manage to extract the board from the background, detect all the transmission lines and measure their size. The size is measured in pixels rather than the actual length.