Learning Topics: Infrared detail enhancement algorithm
learning time: 2020-02-11
Algorithms general introduction:
Using the basic framework of the hierarchical image processing, image filtering using the guide layers.
Effect can be achieved:
A) an infrared image can be significantly compressed to a display range 8bit
b) contrast enhancement
c) increasing detail
d) suppressing noise
- Some conventional algorithm
infrared image Why do we need this algorithm, my previous blog post has been written.
Conventional algorithm:
A) contrast enhancement, representative are CLAHE 1994 classic algorithm
b) an automatic gain control, Gonzalez in its "Digital Image Processing" writes that
c) the histogram equalization HE
D) the Retinex algorithm, can be visible, infrared is not good, especially the zoom noise problems
e) based on hierarchical, 2009 Italian scholars, as I said in this article also detailed.
The algorithm is effective, the problem is the large amount of computation, gradient flipped.
In this paper, based on a hierarchical algorithm, reference 2009 and 2011 Italian scholar Nan Ligong left over papers, more than two article has two advantages: First, reduce the amount of calculation, the second is the gradient will not be reversed.
Algorithm framework as follows:
small
Disclaimer: This article is only high in 2014, Zhu Huanan paper summarizes the study.