The Study Of Measuring Technology For Skin Surface Roughness

The quality of the skins, good or bad, not only reflects the status of human’s physical health, but also has an obvious effect on a person’s external image, which determines the first impressions of people. Therefore, the medical community as well as the industry of skin care products has been working on how to descript the state of the skin quantitatively and objectively. According to the existing theoretical knowledge and references, a measuring system of the degrees of hand skin’s roughness based on the analysis of statistical features of image texture has been studied.First of all, the importance of measuring the state of the skin scientifically and quantitatively both in medical research and the care of beauty and skin is introduced. Besides, the current measure technologies of the skin’s state at home and abroad are summarized briefly. In addition, some kinds of image preprocessing methods are analyzed and conducted with MATLAB. After methods’ results are analyzed and compared with each other, a new method that images are reprocessed by Gabor filter to reduce the noise is proposed so that it benefits the texture images to be analyzed and processed later. Afterwards, some kinds of methods for statistical features related to image texture are researched thoroughly, including the method of central moment, the method of run-length, the method of spatial gray level co-occurrence matrix and the method of local binary modulus. While the actual texture meaning of each feature of the above statistics methods and the shortage of the method of spatial gray level co-occurrence matrix analyzed, a method that combines the method of local binary modulus with the method of spatial gray level co-occurrence matrix is proposed to recognize the features of skin image texture, considering the advantages of the method of spatial gray level co-occurrence matrix.Next, the technologies of pattern recognition and its current development are introduced. An improved BP network algorithm is used as the classification algorithm rather than the traditional ones. Programs aiming to preprocess the skin images and recognize the feathers are developed with MATLAB. What’s more, acquisition system of hand skin images is designed with portable digital microscope and computer.In the end, it proves to be precise and efficient when recognizing the skins’ feathers by the combined method with experiments. And the acquisition system of hand skin images designed in the paper is proved to be theoretically correct and practically possible.