| X-ray security equipments are widely used,especially for transportation ports.In order to find out those prohibited goods more efficiently,users and research staff pay more and more attention to the quality of X-ray security images.Due to the hardware configurations and the multiformity of luggages,luminance changes drasticly between the brightest and the dimmest regions,which means X-ray security images have high dynamic range.Most conventional display technologies are low dynamic range(LDR)in nature.Here comes the problem,how to compress the dynamic range of the X-ray security images in order to be displayed on the common low dynamic range display equipments,without losing any details in the original images.Dual-energy X-ray security equipments contains two kinds of sensors,which produce higher energy X-ray images and lower energy X-ray images respectively.Both of these two kinds of images have their own regions with salient structures.Fusing these images can improve the visual quality.This paper combines a high dynamic range compression method and an image fusion method,and provides a solution to the high dynamic range Dual-energy X-ray security image display problem.In the first part of this paper,the author uses the gradient domain high dynamic range compression method to sovle single energy X-ray security image display problem.Firstly,a Gaussian pyramid should be constructed and calculate the gradients of the images at each level in the pyramid.The gradient attenuation function is deduced from the gradients in the pyramid in a top-down fashion.The attenuation function can penalize larger gradients more heavily than smaller ones,thus compressing drastic luminance changes,while preserving fine details.A new,low dynamic range image is then obtained by solving a Poisson equation.The experimental results show that details in the original images are preserved in the result images.The second part of this paper is to compress the dynamic range of Dual-energy security X-ray images.At the biginning,an energy function is built based on Markov random field fusion modle.Using max-flow/minimum-cut algorithm to minimize the energy function,this fusion model can produce a fusion weight map,directly indicating the fusion regions from different input images.Using this fusion weight map the gradients of the fusion image can be caculated.Then attenuate the gradients to compress the dynamic range of the fusion image.At last,a low dynamic range fusion image is obtained by solving a Poisson equation.The fusion image has both of regions with salient structures,which are from the higher and the lower energe X-ray security images.And the qulity of the fusion image is higher than both of the two compressed single energy X-ray security image. |
