High efficiency video coding (HEVC) is a very popular video coding standard. The HEVC can achieve high coding efficiency with a lower bitrate for intra frame coding. However, it still needs many bits to finish best rate-distortion (R-D) curve. Since there are only 35 directions prediction modes provided in intra prediction module (IPM), HEVC occurs a large distortion when the image contents are out of these prediction directions. In order to obtain a better R-D curve, Zhang et al. [3] recently proposed a simple convolutional neural network (S-CNN) to improve the encoding performance of HEVC. However, S-CNN has to consume more time to encode intra frame coding since it needs to perform more CNN enhancement mode. In order to further speed up S-CNN based intra frame coding, we propose an early termination algorithm to skip CNN. Because the natural images are generally homogenous, we find the mean square errors (MSE) of reconstructed CTU exist high spatial correlation at HEVC encoder. Therefore, a dynamic threshold of MSE is set according to three neighboring encoded CTU blocks to evaluate whether the current reconstructed CTU is useful for the CNN enhancement mode. Simulation results show that the proposed method can achieve faster HEVC encoding process than S-CNN by reducing time increase ratio (TIR) about 12% on an average.

In recent years, automatic license plate recognition (ALPR) system is applied in some traffic-related applications based on deep learning. However, the new ALPR is very difficult to obtain high detection and recognition rates for oblique car license plate (LP). Recently, Silva et al. [5] proposed a warped planar object detection (WPOD) based on deep convolutional neural network (CNN) to overcome the oblique views of LP. Although the WPOD network can achieve the location and rectification of LPs, the loss function of WPOD renders the confidence parameter due to high computational complexity. This also leads to WPOD network cannot locate the optimal LP bounding box. In order to further improve the accuracy of ALPR system, we develop a simple intersection over union (IOU) algorithm to speed up the calculating process of confidence. In this paper, the four-vertex coordinates of the label bounding box and prediction bounding box of oblique LP are used to generate two rectangular boxes, and then a simple IOU algorithm is used to fast calculate the approximate value of IOU. Simulation results show that the proposed ALPR system can arrive a high accuracy of LP recognition about 95.7% on an average. In addition, the proposed system also can achieve higher recognition rate about 1% when compared to the Silva’s ALPR system.

Fingerprint image enhancement is one of the most crucial steps in an automated fingerprint identification system. In this paper, an effective algorithm for fingerprint image quality improvement is proposed. The algorithm consists of two stages. The first stage is decomposing the input fingerprint image into four subbands by applying two-dimensional discrete wavelet transform. At the second stage, the compensated image is produced by adaptively obtaining the compensation coefficient for each subband based on the referred Gaussian template. The experimental results indicated that the compensated image quality was higher than that of the original image. The proposed algorithm can improve the clarity and continuity of ridge structures in a fingerprint image. Therefore, it can achieve higher fingerprint classification rates than related methods can.

In this paper, a fast intra mode decision algorithm for H.264 that exploits the interblock correlation in the intra-mode domain is proposed to reduce computational complexity. Four modes of neighboring coded macroblocks/blocks are considered as the good candidate intra modes of the current block. Experimental results show that the proposed method can efficiently save the computation cost with little degradation in the rate-distortion performance.

An efficient motion-estimation algorithm based on partial block distortion using sorted significant features including bit-plane and mean is proposed. The proposed algorithm can obtain relatively accurate motion vectors with a reduced computational load. Simulation results show that the proposed method achieves its MSE performance very close to the full search method, while requiring only 6-8% of the computation needed by the full search. Furthermore, the performance of our method is better than other algorithms based on partial block distortion search

The H.264/AVC video coding standard uses 7 variable block sizes ranging from 16times16 to 4times4 in interframe coding. The motion estimation with 7 modes needs very high computational complexity. To reduce the complexity of ME module in H.264, we propose a new and fast motion-estimation algorithm based on partial block distortion for sorted significant features including bit-plane and absolute difference of means (ADM). The partial distortion searching (PDS) algorithm, from top-to-bottom line matching scan, is a popular method for fast full search (FSS) in H.264. When the proposed algorithm is combined with the sub-block PDS (proposed-PDS), it can find the same motion vectors as FSS. Furthermore, when the proposed algorithm is combined with the normalized PDS (proposed-NPDS), it can obtain relatively accurate motion vectors with a large reduced computational load. Simulation results show that the proposed-PDS method requires about 90% of the computation needed by the FSS in H.264 without any loss of R-D performance, and the proposed-NPDS requires only about 68% of the computation needed by the FSS with the R-D performance very close to FFS

An efficient fractal image-coding method that exploits the correlation between range blocks is proposed. Four domain blocks mapped by the precious neighboring range blocks are considered as the good candidate blocks of the input range block. Experimental results show that when the proposed method is implemented with a fast fractal coding algorithm, it can further reduce the encoding time and bit rate with insignificant loss of image quality.