Pathology Detection Of Nails Based On Color And Texture Characteristics

The use of medical images in clinical diagnosis and treatment is becoming more common.An important question now worth studying is how to use a large number of images in the image system to assist the doctor in diagnosing the condition.The application of computer image processing in traditional Chinese medicine,the use of intelligent machine to assist doctors to achieve Chinese medicine nail diagnosis is the ultimate goal of this subject.In order to apply the digital information scheme of modern computer image processing technology to the nail pathology diagnosis of Chinese medicine,and to match the classification of nails by computer.In this paper,the analysis of Chinese medicine elephant is applied to modern image processing technology,Algorithm for nail image retrieval.In this paper,we mainly study the nail image detection technique,feature extraction,similarity retrieval and matching in medical images,and propose the nail pathology detection technology based on color and texture.The specific research contents include the following aspects(1)Research on nail image target and background segmentation.In the nail retrieval,it is necessary to extract the target information of the nail in the image,optimize it in the stability of the algorithm and the feature extraction of the nail part,and combine it with the RGB color threshold segmentation algorithm and the light compensation method.This solves the problem of getting nail target and background segmentation.(2)Research on Feature Extraction.In the first experiment,the color and texture features were extracted by HSV color moments and gray level co-occurrence matrices respectively.In the lifting experiment system,the feature extraction based on the gray histogram and the perceptual hash are used respectively.(3)Similarity Matching Research.After the feature extraction,the image processing enters the similar matching stage.In the first experiment,the color and texture features are calculated by Euclidean distance,and the matching result is obtained after Multiple Features fusion.In the lifting experiment,the gray scale histogram feature uses the Euclidean distance to calculate the similarity,the perceptual hash technique uses the Hamming distance to match the similarity,then the Multiple Features fusion is applied to the color image segmentation,that is,the nail image preprocessing,Match the result.(4)A Study on Solving the Problem of Slow Matching.In the process of similarity matching,too many pictures in the image library will cause the match to be too time-consuming,seriously affecting the overall experience of the system.The use of Multiple threading to speed up the matching speed of the retrieval system,the use of Multiple threading in the program can occupy a long time to transfer the task to the background to deal with,thereby improving the efficiency of the entire retrieval system.Based on the color and texture characteristics of the nail pathology detection system can be more intuitive and true expression of nail information and matching characteristics in the medical field has a good application value and application prospects for the diagnosis of Chinese medicine nails to provide a convenient,intuitive,with a strong immersion Sense of visualization applications.