Multi-Window and Line Scan Histogram Features for Bilateral Filtering

Bilateral filtering is the crucial process to enhance the image. This paper aims to improve the bilateral filtering base on the multi-window and line scan histogram scheme. The multi-windows histogram has been introduced to solve the problem when apply to large image by using a number of the window histogram with different weight to estimate the domain filtering function. Anyway, the complexity of this algorithm is increase by multiply of the number of m windows histogram that use for estimating the domain filtering. To improve this, our algorithm that based on the multi-windows histogram is proposed which can reduce the complexity of the filtering from O( mB ) to O( m+B ). Also, our algorithm uses multi-line scan histogram extraction which adapted from dual line scan histogram extraction to reduce the complexity. The experiments show the new algorithm has slightly increase for filtering when increase the number of the window histogram m .