Double-plateaus Histogram Enhancement Algorithm for Low-light-level Night Vision Image
Yang Shubin -, Xi He, Heng Cao, Wanlong Cui · Journal of Convergence Information Technology · 2011
In allusion to characteristics of low contrast, weak-arrangement gray and dim vision in low-lightlevel night vision image, double-plateaus histogram enhancement algorithm is presented. Firstly, median filter is processed on original image histogram to delete those zero-gray values. By selfadaptive setting double-plateaus thresholds as the upper and lower limit for the processed histogram, the algorithm can enhance the image efficiently. By setting a higher threshold, the algorithm can constrain the background and noises. At the same time, it can magnify dim targets and image details by setting a lower threshold. With the proposed algorithm, disadvantages of classical histogram and other plateaus histogram enhancement algorithm are overcome while achieving high contrast. Experiments prove that the proposed algorithm can enhance image contrast effectively and preserve image details simultaneously. Moreover, it can also overcome over-bright phenomenon. It is an effective enhancement algorithm for low-light-level night vision image.