Scalability Tower Multi-scale DR Image Enhancement Algorithm Based on Human Visual Characteristics

Abstract

Objective: Direct digital X-ray (DR) images, which is high resolution, wide dynamic range and rich in a lot of human tissues information, are enhanced to extract a wealth of clinical diagnostic information for early lesions found to provide a good basis for the diagnosis. Therefore, this paper , according to the system for the CCD-DR imaging features, studies a scalability tower multi-scale DR image enhancement algorithm based on human visual characteristics. Method: Firstly, the algorithm uses an improved Laplace Pyramid structure in image decomposition processing. Secondly, the high frequency part can be enhanced by the local nonlinear adaptive contrast enhancement method, the low frequency part of the improved method of histogram equalization combined with human visual characteristics. Lastly, original image will be reconstructed through the repeated extension of image and the results. Conclusion: In this paper, the image enhancement algorithm extends the DR image useful information, highlights the image detail and speeds up image processing speed , while it limits noise amplification and local contrast over enhanced so as to facilitate medical diagnosis and operation.