The Research And Application Of Infrared Image Enhancement Base On MATLAB

Nowadays, to detect and track objects by means of infrared image technologies has become an important sign of modernization, automatization, and intellectualization of military weapons. However, the phenomenon, such as noise pollution and blurred edge caused by natural environment and the inherent characteristics of infrared detectors, may cause mistakes on object identification, while an infrared image with high contrast and abundant details will help identify objects. Thus, in order to reduce such effect and improve image quality, the scouts, according to characteristics of the image, should enhance useful information and restrain disturbing information such as noise of the image, which is the main point of this thesis.This paper will analyze the existing image enhancement processing algorithms of spatial and frequency domain. Experiments have shown that on the condition of MATLAB, the existing image enhancement processing algorithms will also increase the noise when enhancing useful information of the image. This is because such algorithms are often inadaptable to the enhancement processing of different infrared images. To resolve this problem, wavelet analysis technology is thoroughly discussed. As a branch of mathematics derived from Fourier transformation, Wavelet transformation could describe the associated time frequency characteristics of signal by combining spatial and frequency domain. At present, wavelet transformation is widely used in the fields of voice identification, image processing, graphic analysis, communication designing, earthquake forecast, biomedical engineering, and computer vision, in which it has made great achievements.In the paper we will also make researches on the image enhancement method based on wavelet transformation, to which the key is wavelet coefficient enhancement because the sub-selected linear method and non-linear logarithm will also increase the noise. Therefore, several wavelet coefficients are introduced and a wavelet coefficient enhancement method using the pertinence of each scale is also proposed in the paper. According to the results of experiments, the image enhancement algorithm based on wavelet transformation is better both in image enhancement and in de-noise than the traditional image enhancement algorithms.