Views: 1 Author: Site Editor Publish Time: 2020-05-11 Origin: Site
Enhanced infrared thermal image closer to nature
Infrared thermal imaging technology has been more and more widely used in the world. Due to the inherent infrared thermal characteristics of the object and the deformation of the detector, it is difficult to observe and identify the real object in the infrared thermal image. This image is a low-contrast and noisy image, which should be enhanced to make it appear more natural.
The picture shows the original imaging, traditional algorithm, CLAHE algorithm, infrared thermal image comparison of this algorithm
Recently, several enhancement methods for infrared thermography have been proposed. For example, an infrared thermal image enhancement method based on contour wave transform and chaotic particle swarm optimization, infrared thermal image enhancement algorithm based on wavelet transform, infrared thermal image contrast enhancement based on discrete stationary wavelet transform and nonlinear gain operator Algorithms, improved Fourier and Retinex algorithm in discrete cosine domain. But the enhanced quality of these algorithms has a common problem, and their images do not look real and natural.
The picture shows the original imaging, traditional algorithm, CLAHE algorithm, infrared thermal image comparison of this algorithm
In this regard, a novel and simple infrared thermal image enhancement technology is studied, which integrates representative enhancement methods, including: optimized stretching, color conversion, and CLAHE color conversion. Two evaluation methods are considered to verify the performance of the enhancement algorithm. One is the subjective evaluation method, which visually evaluates the enhanced image by using the average opinion score. This indicates that the image of the algorithm is the closest to the human visually. The second is the conventional method. In the second evaluation, the reference image quality, ie the EME value, shows the superiority of the algorithm.
Reference materials:
Sos Agaian, Mehdi Roopaei. Novel Infrared and Thermal Image Enhancement Algorithms. Mobile Multimedia/Image Processing, Security, and Applications. 8755, 2013.
