Image Enhancement via Cloud Cascade Control Based Sub-Image-Clipped Histogram Equalization

Histogram-Based image enhancement techniques attempts to divide the histogram vertically and horizontally to overcome the saturation effect of intensities and over-enhancement. The proposed cloud-based method makes an effort to modify a classic histogram-based image enhancement by concept of cascade control. Two control loops check the quality of the output image by an inner and outer feedback signals and provide optimal vertical and horizontal separating thresholds to control the sub-image-clipped histogram. Feedback loops include of two genetic algorithm mechanisms as two cloud servers with different image quality metrics as their fitness functions. A controller block is also designed to guarantee the convergence of the algorithms. The experimental results, in a bare metal with GPU cloud instances with mass number of processing cores, show the superiority of the proposed method in comparison with traditional image enhancement schemes.