Bit Rate control plays an important role in video coding. Region of Interest (ROI) based rate control has been attracting great attention due to the rapid demands in the region of interest in video coding. The main issue in video coding is the trade off between compression ratio and quality of the reconstructed signal. It is obvious that better quality can be achieved with smaller compression ratio and higher encoded stream bit rate. An optimal coder requires knowledge of the rate distortion (RD) model for the coding scheme. The R-D model is generally built in such a way that the quality of whole frames in a video sequence is taken into account. However in many applications like video monitoring and surveillance, telemedicine, videophone and videoconferencing, some areas in the consecutive frames of the video sequence are more important than others. It is desirable to encode those areas, called region of interest (ROI) with smaller distortion than the rest of the sequence (background). This paper presents a review of the available schemes for bit rate control in region of interest based video coding. Index Terms-Background skipping, bit rate control, content based bit rate allocation, macroblock layer control, ROI video coding, SSIM QP.
The methodology of combining two or more relevant images into a single highly informative image is referred to as image fusion. A new fusion methodology is introduced for combining images obtained from multiple cameras using nonsubsampled shearlet transform (NSST), fuzzy logic and a simple fuzzy neural network (SFNN). The shearlet transform combines the power of multi-scale methods with a unique ability to capture the geometry of multi-dimensional information and is efficient in representing images containing edges. The unique characteristic of shearlets is the utilization of shearing to control directional selectivity, as opposed to rotation utilized by curvelets. The shearlets are not tight edges and therefore it is necessary to perform the synthesis process by iterative methods. A new method, NSST, is introduced for multi-resolution decomposition of input images is introduced. The pixel-based fusion is performed by using fuzzy logic of NSST low-pass coefficients to generate superior quality. The region-based technique is performed by using the SFNN of NSST high-frequency directional coefficients. The SFNN exquisite the set of exemplar input feature vectors and centers a Gaussian function on each remaining one and saves its output label.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.