Near-infrared Detection And Systematic Study Of Breast

Breast cancer is one of common malignant tumour in women, In recent years, the incidence of breast cancer increased year by year, The incidence rate in some cities has risen to the first one . so, Improve the early diagnosis of breast cancer has a very important significance to the treatment of breast cancer. Infrared scanning breast is an important means of diagnosis of breast cancer, it can quickly , painless and Lossless to get the digital of infrared breast images of women. In infrared breast images, the lesions manifested as size, shape irregular tumor shadow, making it difficult to accurately identify the Lesions organizational boundary. And breast image is difficult to reflect the physiological function of breast. Therefore, to extract the effective features of tumor tissue to identify the lesion,easy to make a correct medical diagnosis, the need to suppress image noise and enhance the image in order to improve the infrared breast image quality and detection through image processing and oxygen combine to better detection of breast diseases.In this paper, On the theory of image de-noising, image enhancement and oxygen detection, the main research of the algorithm based on non-orthogonal decomposition adaptive anisotropic Gaussian infrared breast image denoising and Based on morphology, edge detection of infrared breast image enhancement and based on the Dual-wavelength oxygen detection algorithm, the principal tasks are as follows:First, In the infrared breast image denoising algorithm, Introduced a non-orthogonal decomposition algorithm for adaptive anisotropic Gaussian. This method detailed analysis of the direction,scale selection, non-orthogonal decomposition of Anisotropic Gaussian, The simulation, using signal to noise ratio as an indicator proposed for the infrared breast new method for image denoising. Experiments show that the method effectively suppress noise while retaining as much useful detail on the edge of a doctor, in order to provide a more detailed clinical clear message.Second, In some typical breast image is not the issue, carried out image enhancement based on morphology, edge detection. In this paper Proposed three algorithms were enhancing blood vessel method, gray curve shows enhanced method and and ash shadow edge enhanced method. these three methods of threshold parameters and the processing methods were analysised in-depth, especially in the blood vessels of the algorithm taking the vascular morphology-based, machine vision approach to the Vascular image, Very good image shows the breast blood vessels hidden information. From the experimental results obtained, these three methods greatly improved the image of important features of lesion visualization.Finally the use of dual-wavelength oxygen formula, we get the oxygen values of the mammary gland, and carried out in vitro blood model validation, test results show that near-infrared breast cancer detectors can accurately detect the blood oxygen content of the relative changes in the model.In summary, the papers will be the image denoising, enhancement and detection of organic combination of oxygen, while use of medical simulation of breast images as objects, from the many experiments that the algorithm can effectively reduce noise, improve signal to noise ratio to better highlight the image edge information, while enjoying the oxygen content of the value of the mammary gland, in order to raise the value of medical images and diagnostic value. So, for infrared breast image post-processing as an adjunct to medical diagnosis, there can not be ignored significance.