Chi-Kao Chang

An effective and flexible image enhancement algorithm in compressed domain

Images with JPEG format using discrete cosine transform (DCT) is selected for most popular compression standards. Enhancement in the compressed domain offers two major advantages, including low computational complexity and storage space. However, the compression is achieved by block-based transform; therefore, it is hard to enhance image globally. In this paper, the main issue is to develop a global enhancement method that effectively reduces the introduced blocking artifacts and achieves excellent visual quality of enhancement. We propose a combined DCT matrix representation, which consists of 8n × 8n pixel arrays, to enhance the global information on images for removing block artifacts in compressed-domain. In addition, we propose the multi-enhancement factors based on spatial frequency for image enhancement in compressed domain. From the simulation results, the proposed method achieves not only excellent improvement for image enhancement but also reducing the blocking artifacts significantly.

Robust Player Tracking and Motion Trajectory Refinement for Broadcast Tennis Videos

Player tracking and trajectory detection play an important role in the content analysis of broadcast tennis videos. It is still a challenge because the player size is small and many noises and interference exist in a tennis court, and therefore often results in a detection failure. In this paper, we propose a robust technique for player tracking and trajectory modification using an adaptive Kalman filtering. Experimental results indicate that the proposed method improves the successful rate of player tracking significantly, especially for the upper court players.

Constructing a discriminative visual vocabulary with macro and micro sense of visual words

Visual vocabulary representation approach has been successfully applied to many multimedia and vision applications, including visual recognition, image retrieval, and scene modeling/categorization. The idea behind the visual vocabulary representation is that an image can be represented by visual words, a collection of local features of images. In this work, we will develop a new scheme for the construction of visual vocabulary based on the analysis of visual word contents. By considering the content homogeneity of visual words, we design a visual vocabulary which contains macro-sense and micro-sense visual words. The two types of visual words are appropriately further combined to describe an image effectively. We also apply the visual vocabulary to construct image retrieving and categorization systems. The performance evaluation for the two systems indicates that the proposed visual vocabulary achieves promising results.

A Novel Texture Description for Liver Fibrosis Identification

In this study, the proposed texture description method is applied to obtain the description of ultrasound images of hepatic parenchyma. The result of performance characteristics for distinguishing liver fibrosis and normal liver is shown. The diagnostic performance is accessed on two different approaches and two set of parameters including CO-LBP 50 × 50, CO-RLBP 50 × 50, CO-LBP 75 × 75 and CO-RLBP 75 × 75. We find that CO-RLBP method is better than that of CO-LBP method in overall accuracy.

Image Categorization Using Macro and Micro Sense Visual Vocabulary

Visual vocabulary representation approach has been successfully applied to many multimedia and vision applications, including visual recognition, image retrieval, and scene modeling/categorization. The idea behind the visual vocabulary representation is that an image can be represented by visual words, a collection of local features of images. In this work, we will develop a new scheme for the construction of visual vocabulary based on the analysis of visual word contents. By considering the content homogeneity of visual words, we design a visual vocabulary which contains macro-sense and micro-sense visual words. The two types of visual words are appropriately further combined to describe an image effectively. We also apply the visual vocabulary to construct image categorization system. The performance evaluation for the system indicates that the proposed visual vocabulary achieves promising results.

Computation-Aware Motion Estimation Based on Search-Point Constraint

Computation-aware (CA) concept has been proposed to efficiently allocate computational resources under enviroments of power constraint. However, the computations of motion estimation could be overloaded in limited computing condition, even using the CA algorithms. In this paper, we introduce a method that achieves motion estimation in the extremely insufficient computational budget. We proposed a modified PCDB list to achieve better visual quality. To decrease the search points requirement, subsampling technique is combined with the proposed algorithm. The proposed method significantly reduces the requirement of search points in computing motion vectors. Satisfactory visual quality and computational consumption can be achieved under the extreme condition.

Subsampling Based Computation-Aware Motion Estimation

Computation-aware (CA) scheme has been proposed to efficiently allocate computational resources. However, the computations of motion estimation could be overloaded in limited computing environments, even using the CA algorithms. In this paper, we introduce a method that achieves motion estimation in the extremely insufficient computation power. We proposed a low-distortion based PCDB list to achieve better visual quality. To decrease the computation-power requirement, sub sampling technique is coupled with the proposed algorithm. The proposed method significantly reduces the requirement of computation power in computing motion vectors. Satisfactory visual quality can be obtained under the extreme condition.

A DCT-Manipulation Based Algorithm for Scalable Video Coding

Scalable video coding (SVC) techniques, which reuse the existing motion vectors, provide a way to convert the compressed information between the delivery systems and terminal devices. In this paper, we aim on the increasing of the visual quality for video downscaling in DCT domain. For arbitrary video downsizing, we propose an efficient algorithm, which manipulates compressed video in the compressed domain, for arbitrary video downsizing. First, the original 8(8 block size is properly enlarged according to the downscaling factor. After the downscaling procedure, the downscaled image blocks are reconstructed to 8(8 matrix to comply with the JPEG format. Finally, we introduce a small-range searching method to increase the accuracy of motion estimation, so that we can approximate the best motion vector. The experimental results indicate that the proposed method provides sufficient improvement for arbitrary video downscaling in the compressed domain.