Research On Histogram Equalization Image Enhancement Method Based On Measure Of Medium Truth Degree

Image enhancement strengthens the useful information which is collected in the original imageto meet the requirements of post-image processing. Histogram equalization is not only the classicalmethod of image enhancement, but also one of the most commonly used methods in imagepreliminary processing.In this paper, a traditional histogram equalization method and a retained gray level histogramequalization method have been studied in depth and the advantages and disadvantages of the twomethods have been analysed. The traditional histogram equalization may be merged excessively sothat the processed image may lose details thus leads to a failure in reaching better enhancementeffect. Retaining high or low gray level histogram equalization can overcome this drawback, but anexact determination method has not been reached in the choice of retaining the high gradation or thelow gray level. To address this issue, using the level grayscale pixel distribution characteristics andthe measure of medium true degree, this paper determines the relevant judgment function andproposes the reserved gray-level histogram equalization method based on the measure of mediumtrue degree. Experimental results show that the method can be a good judgment in choosing toretain high grayscale or low grayscale so as to achieve a better image enhancement effect.The improved histogram equalization has really improved a lot in treatment effect, but becauseit is implemented in software, the processing speed is not fast. In order to solve this problem, thispaper designs a FPGA hardware module to run the image enhancement algorithms to impro veprocessing speed. Experimental results show that in the premise of not changing the imageprocessing effect realized by software, the implementation of FPGA hardware greatly improves theimage processing speed. So FPGA hardware implementation is more applicable in engineeringpractice.