Infrared Image Enhancement Using Nearest Neighborhood Contrast Pixel Matching

The infrared images are low contrast images and
due to the imaging environment, the details in the image
may appear fuzzy. So, in order to address this problem
and also to provide images which is noise-free has been
proposed an infrared image enhancement technique
dependent on nearest neighborhood contrast pixel
matching. In which the original image is pre-processed by
filtering the noise in the image. Secondly, the edge
information is extracted without being lost by fractional
edge detection. Saliency weight mapping is done to avoid
the loss of sensitive pixels. Then, by applying the
algorithm, initial pixel contrast mapping is done. Finally,
the image is reconstructed to obtain the enhanced image.
The analysis has higher average value of peak signal-tonoise ratio compared to other approaches. Moreover, the
enhanced results show that the images have been enhanced
with higher quality and noise-free indicating our method
achieves better performance than other methods .