| With the rapid development of medical image processing in the field of compute vision,the computer-aided diagnosis has become a research hotspot.The chest X-ray is an important basis for diagnosing the diseases of the lungs, bones,heart and other parts for doctors. Due to the complexity of the structure of human and X-rayscattering and equipment noise, the quality of the X-ray chest image declined. Therefore, TheX-ray chest image needs processing for laying the foundation to provide reliable diagnostic,including: enhancement to improve image quality, segmentation to highlight the region ofinterest, recognition of lung nodules to check the lesions.This paper firstly describes some basic methods of medical image enhancement,including contrast enhancement and edge detection strengthens, and gives some briefdescription. To study the image enhancement method based on fuzzy set theory, a generalizedfuzzy enhancement algorithm with adaptive threshold has been proposed on the basis ofprevious work. It can be used to detect edge and noise of the business as a standard deviationimage enhancement of quality assessment standards, automatic selection fuzzy parameter,realize the generalized fuzzy enhance the image to be automatic optimization. And thenexpounds the chest radiograph, based on the theory of fuzzy image segmentation methods,including classical k-means clustering algorithm, C-mean clustering algorithms, and based onthese two methods of improving methods, analyze their advantages and disadvantages andgives a lung contour segmentation image. Finally, fuzzy pattern recognition was applied tolung nodules detection and identification, according to the characteristics of benign andmalignant pulmonary nodules extracted, the application of maximum membership degreeprinciple towards the identification of lung nodules of benign and malignant research, andgiven the specific structure and process. Experiments show that this algorithm can achieve a better recognition result. |
PDF Full Text Request