S-LIGHT: Synthetic Dataset for the Separation of Diffuse and Specular Reflection Images

Several studies in computer vision have examined specular removal, which is crucial for object detection and recognition. This research has traditionally been divided into two tasks: specular highlight removal, which focuses on removing specular highlights on object surfaces, and reflection removal, which deals with specular reflections occurring on glass surfaces. In reality, however, both types of specular effects often coexist, making it a fundamental challenge that has not been adequately addressed. Recognizing the necessity of integrating specular components handled in both tasks, we constructed a specular-light (S-Light) DB for training single-image-based deep learning models. Moreover, considering the absence of benchmark datasets for quantitative evaluation, the multi-scale normalized cross correlation (MS-NCC) metric, which considers the correlation between specular and diffuse components, was introduced to assess the learning outcomes.

Variational Model for Single-Image Reflection Suppression Based on Multiscale Thresholding

Reflections often cause degradation in image quality for pictures taken through glass medium. Removing the undesired reflections is becoming increasingly important. For human vision, it can produce much more pleasing results for multimedia applications. For machine vision, it can benefit various applications such as image segmentation and classification. Reflection removal is itself a highly illposed inverse problem that is very difficult to solve, especially for a single input image. Existing methods mainly rely on various prior information and assumptions to alleviate the ill-posedness. In this paper, we design a variational model based on multiscale hard thresholding to both effectively and efficiently suppress image reflections. A direct solver using the discrete cosine transform for implementing the proposed variational model is also provided. Both synthetic and real glass images are used in the numerical experiments to compare the performance of the proposed algorithm with other representative algorithms. The experimental results show the superiority of our algorithm over the previous ones.