A new image enhancement process based on an integrated neighborhood
dependent nonlinear approach for color images captured in various
environments such extremely low lighting, fog or underwater. The new
process is a combination of two independent processes: luminance
enhancement and contrast enhancement. The luminance enhancement, also
regarded as a process of dynamic range compression, is essentially an
intensity transformation based on a specifically designed nonlinear
transfer function, which can largely increase the luminance for dark
regions of the image but only slightly change the luminance for the
bright regions of the image. The contrast enhancement transforms each
pixel's intensity based on the relationship between the pixel and its
surrounding pixels. The output of the contrast enhancement is a power
function of the luminance of the input image. The exponent of the power
function is determined by the information obtained in the original
luminance image, which is the ratio between the result of neighborhood
averaging and the luminance of the center pixel. After contrast
enhancement, the luminance image is converted back to a color image
through a linear color restoration process with color saturation and hue
adjustment.