Least-Squares-Based Switching Structure for Lossless Image Coding

Abstract

Abstract-Many coding methods are more efficient with some images than others. In particular, run-length coding is very useful for coding areas of little changes. Adaptive predictive coding achieves high coding efficiency for fast changing areas like edges. In this paper, we propose a switching coding scheme that will combine the advantages of both run-length and adaptive linear predictive coding. For pixels in slowly varying areas, run-length coding is used; otherwise least-squares (LS)-adaptive predictive coding is used. Instead of performing LS adaptation in a pixel-by-pixel manner, we adapt the predictor coefficients only when an edge is detected so that the computational complexity can be significantly reduced. For this, we use a simple yet effective edge detector using only causal pixels. This way, the proposed system can look ahead to determine if the coding pixel is around an edge and initiate the LS adaptation in advance to prevent the occurrence of a large prediction error. With the proposed switching structure, very good prediction results can be obtained in both slowly varying areas and pixels around boundaries. Furthermore, only causal pixels are used for estimating the coding pixels in the proposed encoder; no additional side information needs to be transmitted. Extensive experiments as well as comparisons to existing state-of-the-art predictors and coders will be given to demonstrate its usefulness.

A Weighted and Combined Super Resolution Reconstruction Algorithm

“…Specially, if 2 p = , then L p norm estimators become L 2 norm estimator [4], [5], [8], [9], and formulation (2) is the least squares formulation [4], [8], [9]. If 1 p = , then L p norm estimators become L 1 norm estimator [2], [5].…”

“…When the system noise is white additive Gaussian noise, the least squares approach will result in the ML estimation [2], [8]. However, if the system noise is not Gaussian noise, the least square estimation will become a non-robust mean estimation and result in a HR image with visually apparent errors.…”

“…Many SR reconstruction methods [2], [4], [5], [8], [9] are based on L 1 or L 2 norm estimator. When data fidelity term is measured by L 2 norm it constructs the least square method, which is computationally easy, but is sensitive to data outliers and HVS is not optimal in the sense of least square.…”

A fuzzy neural network based adaptive predictor with P-controller compensation for lossless compression of images

“…After that, the prediction process is repeated again but with the component B k u in existence this time (denoted as Type 2). Moreover, the experiment can be classified into two RALP [3] cases, one with no online training area (denoted as Case 1), and the other with the predefined four-pixel online training region (denoted as Case 2). In Table I, the results obtained by setting the B matrix to be a zero vector (Type 1) are shown in the columns denoted as A1 and A2, and that of obtained by using the proposed "Pcontroller" compensator (Type 2) are shown in the columns denoted as B1 and B2 respectively.…”

“…Table II gives actual bit rates by JPEG-LS [2], CALIC [1], RALP [3] and TMW [4] for a set of fourteen test images. The results of JPEG-LS, CALIC, RALP and TMW are taken directly from [3].…”