AN ALGORITHM FOR DETECTING STAINS ON MASK SURFACE

To accurately identify the residual stains on the surface of masks, a new algorithm based on digital image processing technology was proposed. Firstly, the mask image is processed by illumination and denoising, and the contrast of the mask image is improved by image enhancement. Then, the mask image is segmented by the threshold iteration method, and the preprocessed image is obtained. To compare the effect of different edge detection operators on mask stain recognition, six commonly used operators are used to detect the edge of the preprocessed image, and then the connected regions are marked by the connected region algorithm. Finally, the connected domain is used to determine whether there are residual stains on the surface of the mask, and the results judged as stains are used for image location. The experimental results show that, the algorithm in this paper does not need to train the positive and negative samples in advance, but it can obtain roughly the same accuracy as the neural network method.