Research On Photon Counting Image Fusion And Colorization Technology

With the development of low-light-level night image processing technology,the photon counting image fusion and colorization technology have become a research hotspot in recent years.The photon counting image fusion and colorization technology have shown excellent application scenarios in aerospace,military operations,military and civil scene monitoring,because they can extend the perception of human visuals.In order to obtain plentiful information and natural color of photon counting images,this paper performs image fusion and colorization on photon counting images.Firstly,the MPPC(Multi-piexl Photon Counter)photon counting imaging experimental platform was designed and built,and the software running program of the optical experiment platform also was written.Under different illumination conditions,the measured target images were scanned by the MPPC detector and it was completed that acquisition work of the measured target images.And the photon counting source image was obtained,which can support the development of the subsequent research content of this paper.Secondly,the image was obtained by the image fusion algorithm with regional energy fusion rules,but the fusion image had an unclear outline,severe noise,and the regional energy fusion algorithm had a long time-consuming operation.So,the image fusion algorithm with improved regional energy fusion rules was proposed in this paper.In high-frequency part of the two source images,the energy of the corresponding region,neighborhood mean square,matching degree and threshold deviation are calculated.And the threshold is determined by pixel values and mean values of regional image.If the matching degree is greater than the threshold value,weighted regional energy fusion method is used to solve it.Otherwise,the enhanced pixel value of source image with larger energy in local region is selected as pixel value in corresponding fusion image.Using the image fusion algorithm with improved regional energy fusion rules,it was proved that the details of photon counting fusion image were more apparent,and target was more accessible to recognize,which verified the effectiveness of the algorithm in this paper.Finally,since the photon counting image is gray scale image,and the human eye recognition of the gray scale image is significantly lower than that of the color image.In the existing color transfer algorithms,there were the color misinformation and color unnatural,and the algorithms run for long time.Therefore,the improved color transfer algorithm of luminance remapping with multi-parameters was proposed in this paper.In this algorithm,the parameters which depend on the number of columns and luminance values of the colorreference image and the gray scale target image were introduced.On the basis of the multi-parameters the luminance values of the color reference image were calculated to ensure the color reference image and the gray scale target image had the same range of luminance values.According to the correlation of the neighborhood,each pixel of gray scale target image was matched with the pixel of the color reference image.The simulation results showed that the occurrence of color misinformation was decreased in color transfer method of luminance remapping with multi-parameters.At the same time,the colorized photon counting image was natural in color,which improved the recognition rate of the target in the photon counting image by the human eye,and enhanced the understanding of the photon counting image scene.In this paper,the program running time of the improved image fusion algorithm and the improved color transfer algorithm were reduced effectively.In the improved image fusion algorithm,the values of information entropy,average gradient and spatial frequency of are about 20%,25% and 30% higher than that of fusion algorithm.And in the improved color transfer algorithm,the objective evaluation index is also improved respectively.