Research On The Key Techniques Of Infrared Image Enhancement

Histogram equalization is a classic algorithm in image enhancement.Because of its simple operation and low complexity,it is of value and can be used for real-time enhancement of video.Although the traditional histogram equalization can improve the image contrast,excessive enhancement and loss of details also exist.Therefore,the improvement of the histogram equalization algorithm has been a research hotspot in the field.After analyzing the shortcomings of the existing improved histogram equalization algorithms,this thesis proposes an infrared image contrast enhancement algorithm based on histogram dual control,and realizes the engineering application of infrared video contrast real-time enhancement.Meanwhile,it uses the sub-block Discrete Cosine Transform to enhance the local contrast of the infrared image.The main work of this thesis is as follows:(1)A platform histogram equalization algorithm based on Gamma correction is proposed.First,the platform threshold of the histogram and the Gamma correction parameter value are calculated through the proportion of the effective gray levels in the histogram.Using the two parameters obtained above,the mean normalized histogram is processed for large data clamping and small data Gamma correction.Finally,postprocessing and histogram equalization are performed.(2)A histogram equalization algorithm based on double Gamma correction is proposed.First,the number of effective gray levels of the pixel points of the input image is calculated.Then,the gray level ratio in the histogram data is normalized by the mean value to obtain two Gamma correction parameters.For the large data part and the small data part of the mean normalized histogram,the Gamma correction operation of the histogram data is performed respectively.Finally,post-processing and histogram equalization are performed.(3)A local contrast enhancement algorithm of infrared image based on Discrete Cosine Transform is proposed.First,the image is subjected to sub-block Discrete Cosine Transform.Each Discrete Cosine Transform coefficient is enhanced by ring nonlinear Gaussian weights.Finally,Inverse Discrete Cosine Transform is performed on each sub-block to obtain an infrared image with enhanced local contrast.(4)A real-time infrared video enhancement system was developed on the Qt platform based on an infrared continuous zoom lens,movement and data collector.The system is developed based on Qt5.8.0 platform and GUI interface design.Capable of real-time processing of captured 720*576 infrared video at 25 frames per second,which can effectively enhance infrared video in different scenes.The system is easy to operate,has not need to adjust parameters manually,can be enhanced in real time,and is practical.