Data Mining Based On Point Set And Image Denoising Based On Total Variation

Skin detect is widely used in image process, all the currently existing skinmodels have ?aws, this article mainly uses a new filtering method and supportvector machine (SVM) to perform data mining with the point set of skin color,get a new skin detect model, and then skin areas in images can be detectedmore accurately. The new obtained model will be compared with existingmodels in di?erent color spaces.Image is an important way for people to get information. However, be-cause of various negative factors'e?ects which come from the process of imagegeneration and transmission, the quality of image is usually degenerated. Inthe meantime, it might be under noise pollution as well. In recent years,images denoising via TV regularization causes widely concern, but researchdiscovers that it still has some ?aws. For example, it can't solve the con?ictbetween image detail restoration and noise suppression. This article importsBregman distance on the basis of total variation, uses iterative method toperform images denoising, and improve original methods. Finally, this articlecompares the improved model with original model through several examples.