The Feature Extraction Of Diabetes Based On Non-Fluorescence Fundus Image

With computer hardware and software and the rapid development of the national economy, computer image processing technology is playing a more and more important role in the national economy. Digital image process technology with a high rate of accuracy, good reproducibility and high processing speed can be considered more and more seriously. Particularly in the medical field, computer image processing technology often subjective judgment of the doctor into a qualitative approach things from the computer set artificial basin side, and thus enhance stability and objectivity of the judge to reduce the labor intensity, this would be need to set some parameters for the calculation method. Fundus images lesions disease diagnosis and research of high value, so it is very important for the extraction.Fundus medical image processing technologies in domestic and foreign development has matured, technology has gradually improved. Fluorescein angiography is the current technology used in the diagnosis of diseases hospitals, means existing technologies are basically the Fluoresce in angiography images for image processing on the basis of results MR diagnosis of the disease and is so out of the analysis and judgment. However, the use of fluorescein angiography has higher costs, even will cause a lot of adverse reaction fluorescence, such as heart palpitations, nausea, vomiting, collapse response, a body rash, liquid leakage caused pain. So Fluoresce in angiography images to the use of non-disease treatment will bring convenience to ensure the safety of the patients, meanwhile the cost can be reduced.With the emergence and development of digital image processing technology, computer analysis of fundus images become into pathologic diagnosis and the necessary aids. But the fundus image analysis technique is still ineffective there are deficiencies. For qualitative analysis to the traditional manual way of entering the highly automated quantitative analysis of the computer, fluorescein angiography images to avoid the high costs and the adverse impact on patients. The artificial retina is achieved with a small number of images involved in the extraction of diabetic lesions accurately. The increase in new clinical ophthalmologists and theoretical research means Third Eye Hospital of Jilin University in Changchun City and non-cooperation fluorescein angiography image analysis and research Experimental data provided by the hospital (fundus images), Fundus image processing algorithm for in-depth study. Based on the study of development in a non-Fluoresce in angiography images of diabetes diagnosis, the system can be used for diagnosis of diabetes and for the quantitative analysis of lesion extracted clear diabetic lesions doctors can help patients with objective and accurate access to the actual situation, in order to make a diagnosis. To effectively deal with a computer analysis of the fundus images, it is essential to fully understand the characteristics of diabetes in the retina. Only in this way can the true medicine and computer image processing technology combine to develop a system for medical diagnosis.Scanners and other equipment used for image input to the computer and digital image preprocessing, fundus diabetes lesions on computers extracted from the lesion to quantify get a doctor with the various parameters required by hardware output seen results. As fundus images are very complex, most t raditional image processing metlhods are not well resolved in the enhancement of fundus images, image segmentation, quantify lesion detection, and other aspects. It needs to develop suitable fundus image processing algorithms. Fluorescein angiography can bring about real image analysis and research. Based on this study, it is to determine the scope of the large amount of information and access on the basis of advice from medical experts in light of reality, fluorescein angiography of the non-image processing technology for an in-depth study:(1) Non-Fluorescein angiography images of the retina image processing are general pretreatment with 256 color bitmap. In this paper, it is 24-bit True Color Bitmap-Fluorescein angiography images. Image is usually the time of the acquisition, equipment, illumination, and the impact on the surrounding environment, to a certain extent, there is noise, affect image quality, we need to enhance the image filtering and other processing to enhance image quality, Image preprocessing to lay a good foundation for the follow-up work. Light color fundus images, L~* a~* b~* color using the RGB color model to replace the model approach to more accurately detecting color fundus images of the aims of this method is not only applicable to the detection of color fundus images, Color image processing and provide for all the new ideas. Propose a model based on the color filter based on the color space vector adaptive median filtering method better able to filter noise and improve image quality. Fundus images combine the features of the fundus pseudo-color image enhancement methods and approach Fundus enhanced image interpretation is more favorable to doctors for follow-up treatment fundus images provide a good guarantee.(2) Diabetes various periods (like a period of 2) lesions noticed lesions (hemorrhage, glass membrane warts, hemangioma) This paper proposes a threshold-based segmentation method, the extraction of diabetes doctors noticed lesions can best serve the needs of According to the non-diabetic fundus fluorescein angiography images show features For some of the weaknesses of the appropriate algorithm to improve and accelerate the speed of image segmentation, the lesion to extract more dear, more accessible doctors in the diagnosis of diabetes.(3) For this non-fluorescein angiography image processing algorithms based on the study a non-fluorescein angiography in the diagnosis of diabetes Image Design and Implementation.This paper analyzes and researches fluorescein angiography images, according to the existing image-processing techniques, collaborated with the improved image processing algorithm for image analysis and processing, extraction clear focus images doctors who will be able to make a clear judgment based on extraction of the tumor was represented by the disease. In order to make a clear judgment on diabetes, fluorescein angiography images ultimately achieve the same result. This paper presents a view of the non-diabetic fluorescein angiography images from the lesions, get a clear image of the lesion. The method of image segmentation technology allows ordinary diabetic lesions fundus images to extract good results in order to facilitate the diagnosis and treatment of diabetes.This article will discuss in detail the above-fluorescein angiography image processing and realization, also conduct in-depth problem analysis and study in the realization of the process.