HDR Enhancement Based on Bi-curve Mapping for Image Sequences

In this study, a video enhancement method based on high dynamic range processing was proposed. This method can enhance images that have backlighting problems and are underexposed or overexposed to improve visibility. The proposed method comprises pre-processing, illumination judgment, tone mapping, and color restoration. Initially, a small segment of an image is used to calculate an average luminance value. Next, the average luminance of each segment of the image and the reference average previously calculated are used to judge and classify the illumination of the image to facilitate the subsequent selection and setting of a tone mapping curve. Finally, these classification results are used for tone mapping to restore color and increase visibility. The main feature of this method is that two equations are used to form an S-curve function for enhancing images. This advantage enables new tone mapping curves to replace conventional exponential and logarithmic operations. Therefore, computational complexity can be substantially reduced, rendering this method appropriate for real-time digital video recording systems.