Local Histogram Equalization Algorithm And Its Applications

This paper focuses on the study of Local Area Histogram Equalization (LAHE) and its applications in pneumoconiosis X-ray films enhancements. Pneumoconiosis is one of the major occupational diseases of which no cures have turned out effective so far. X-ray film examination is necessarily adopted to diagnose and discern pneumoconiosis. However, X-ray film features an excessively broad range, over -exhaustive details and poor contrast. Also, it is not obvious as to the pathological changes of early pneumoconiosis symptoms. These factors add difficulty to the early diagnoses of pneumoconiosis. With the help of digital image processing, the contrast of pneumoconiosis X-ray film can be enhanced, and such early symptoms will be emphasized with better clarity in the feedback. Thereby, we are hoping to reduce the misdiagnosis rate.Histogram Equalization (HE) as an image contrast enhancement technique is widely used. By improving the visual effect of images considerably, it assists the viewer in discovering hidden messages. However, satisfaction is hard to be achieved by HE when we are interested in the information presented in smaller areas with relatively even grayscales. Local Area Histogram, as an extension based on standard HE technique, has given pleasing results. However, LAHE shows a lack of flexibility because there is only one parameter - size of the rectangle window. Small windows help increase the detail clarity of image, but weaken the overall outline of the image. Big windows, on the other hand, give better outline by sacrificing the detail clarity. In this paper, the author tries to improve LAHE from two aspects. On the one hand, we can modify the structure of the local histogram; on the other hand, we can also adjust the size of rectangle window adaptively. At last, the results using LAHE and its modifications will be given and comparisons among these different algorithms are to present as well.