Abstract
The estimation of image quality and noise perception still remains an important issue in various image processing applications. It has also become a hot topic in the field of photo-realistic computer graphics where noise is inherent in the calculation process. Unlike natural-scene images, however, a reference image is not available for computer-generated images. Thus, classic methods to assess noise quantity and stopping criterion during the rendering process are not usable. This is particularly important in the case of global illumination methods based on stochastic techniques: They provide photo-realistic images which are, however, corrupted by stochastic noise. This noise can be reduced by increasing the number of paths, as proved by Monte Carlo theory, but the problem of finding the right number of paths that are required in order to ensure that human observers cannot perceive any noise is still open. Until now, the features taking part in the human evaluation of image quality and the remaining perceived noise are not precisely known. Synthetic image generation tends to be very expensive and the produced datasets are high-dimensional datasets. In that case, finding a stopping criterion using a learning framework is a challenging task. In this paper, a new method for characterizing computational noise for computer generated images is presented. The noise is represented by the entropy of the singular value decomposition of each block composing an image. These Singular Value Decomposition (SVD)-entropy values are then used as input to a recurrent neural network architecture model in order to extract image noise and in predicting a visual convergence threshold of different parts of any image. Thus a new no-reference image quality assessment is proposed using the relation between SVD-Entropy and perceptual quality, based on a sequence of distorted images. Experiments show that the proposed method, compared with experimental psycho-visual scores, demonstrates a good consistency between these scores and stopping criterion measures that we obtain.
Highlights
Realistic image computation mimics the natural process of acquiring pictures by simulating the physical interactions of light between all the objects, lights and cameras lying within a modelled 3D scene
These simulations are intrinsically subject to visual noise, which degrades the quality of the images obtained, due to their use of calculation methods based on Monte Carlo integration
The noise reduction requires an increase in the number of light path samples to be used, without a reliable criterion being available at the moment to define a threshold for stopping the calculations
Summary
Realistic image computation mimics the natural process of acquiring pictures by simulating the physical interactions of light between all the objects, lights and cameras lying within a modelled 3D scene. Sampling is performed through the build of random light paths between the camera and the light sources lying in the 3D scene: a ray is sent from the camera location through a pixel and is randomly reflected by the surface of the first object it encounters. The process is recursively performed until a light is reached or a Russian roulette process decides to stop the path. Numerous light paths are build per pixel and according to the law of large number and the Monte Carlo approach, the mean of the samples for each pixel converges to the solution following a
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
