Research On Multi Exposure Image Fusion Algorithm In Multi Type Scenes

In the era of rapid development of science and technology,there are many application scenarios in the field of image processing.Aiming at the problem that the acquisition defects of hardware devices make it impossible to obtain the full dynamic range of the real scene,multi exposure image fusion technology arises at the historic moment.Nevertheless,multi exposure image fusion technology have a series of problems,such as the loss of texture details,halo artifacts and so on.Therefore,multi exposure image fusion technology has important research value.In order to improve the detail information of multi exposure fusion algorithm in static scene and ghost removal in dynamic scene,this study offers a method of calculation on many different types of fields.First of all,an adaptive multi exposure image fusion method is proposed based on weighted least squares is proposed for static scenes.Second of all,an improved intensity mapping function for multi scale exposure image fusion is proposed for dynamic scenes.And in-depth research has been carried out in the following aspects:Firstly,to solve the problem of image edge detail degradation caused by traditional weight factors,an adaptive multi exposure image fusion algorithm based on weighted least squares method is proposed.In order to generate adaptive weights efficiently,the weight based on pixel intensity is calculated according to the overall brightness and the importance of adjacent exposure images.Then,the pixel value based on the global gradient is calculated according to the importance of the pixel value in the range of relatively large global gradient compared with other exposures.At the same time,aiming at the problem that the existing algorithms can not guarantee the image details of the brightest and darkest regions,this study uses a framework based on weighted least squares,and completes the multi exposure image fusion algorithm by combining with adaptive weights.In the first place,the weighted average of gradient global exposure is used to construct the vector field.In another,the fine details of the vector field are calculated by using the weighted least squares optimization framework.What's more,the final fusion image is synthesized by using the adaptive weight map and the extracted details.Experimental results show that the proposed algorithm has a better processing effect on image color and local texture details.Secondly,aiming at the problem of ghost detection and elimination in multi exposure image fusion,a multi scale exposure image fusion algorithm based on improved intensity mapping function is proposed.Above all,the high and low contrast of the input frame is determined based on the reference frame image.Furthermore,the high contrast region is used for structure consistency detection to get the ghost region.Then,the improved energy function based on intensity mapping relationship is used to further detect the ghost information in the low contrast region.Last but not least,the multi scale block matching algorithm is used for fast fusion and the resulting image is generated.Experimental results show that the proposed method can effectively remove the ghost and retain the color and detail information of the image.Thirdly,according to the algorithm proposed in this study,a multi exposure image fusion application system is developed.The system encapsulates and visualizes the multi exposure image fusion algorithm in multi type scenes.Users click different algorithm tabs to upload multi exposure image sequences in real time.The system processes the uploaded results quickly and displays them on the interface.At the same time,the user clicks the image quality evaluation tab to upload the required algorithm results.The system makes an objective analysis of the image quality evaluation and displays the evaluation value.