High Dynamic Range Image Deghosting And Color Processing

Modern digital camera which using single exposure can not cover the entire dynamic range of natural scenes.The low dynamic range image in traditional exposure and less exposure area cannot contain all the details.The high dynamic range image technology uses series of low dynamic range images with different exposure to expand the coverage of single image and enhance image details.High dynamic range technology based on its good visual experience has been gradually applied in geographic information systems,medical imaging,video surveillance and video effects and other fields.All in all,high dynamic image technology has a very important theoretical research value and extensive commercial applications.The existing fusion technology to ensure the fusion image is obtained by special equipment which requires that the background of different exposure images to be fused must be completely still.However,images getting in the real scene mostly contain moving objects such as dynamic elements.This will cause the fused high dynamic range image contains ghosting.The existing ghost removal technology usually remove all moving objects directly in the fusion result,which makes the high dynamic range image loss of original image fidelity to a certain extent.To improve the multi exposure image fusion technology based on gradient and brightness,this paper adds saturation as one of the measure factors,and then use Gauss function as the weight transfer function instead of the original triangle symmetric function.The improved method can retain more pixels exposed well in the fusion result.As a result,this can improve the quality of the fusion results.This method uses the direct fusion of multi exposure images,using appropriate measure factor to separate the static background and moving objects.The algorithm firstly selects low dynamic range of different exposure images.In order to pretreatment before the background modeling,histogram equalization is applied on these images to eliminate the effects of illumination and other external factors with the original image.The use of background modeling are compared with the residual images to detect the position of the moving object.Then by setting different fusion weights of moving objects and static background,weakened the body to shift animal ghost interference we can obtain the final fusion image.High dynamic range image without ghosting effects obtained by this typical algorithm has high quality.This paper designs initial algorithm based on its feasibility is proved by the Matlab simulation and test image sequences validate the subjective experimental experimental effect.The results show that this algorithm can effectively synthesize high dynamic range without ghosting image.In order to improve the applicability of the algorithm in this paper,we can retain a single moving object in high dynamic range image fusion in the final results or keep the trajectory of moving objects by modifying the weights of graph.