Sangkeun Lee

An Efficient Content-Based Image Enhancement in the Compressed Domain Using Retinex Theory

The object of this paper is to present a simple and efficient algorithm for dynamic range compression and contrast enhancement of digital images under the noisy environment in the compressed domain. First, an image is separated into illumination and reflectance components. Next, the illumination component is manipulated adaptively for image dynamics by using a new content measure. Then, the reflectance component based on the measure of the spectral contents of the image is manipulated for image contrast. The spectral content measure is computed from the energy distribution across different spectral bands in a discrete cosine transform (DCT) block. The proposed approach also introduces a simple scheme for estimating and reducing noise information directly in the DCT domain. The main advantage of the proposed algorithm enhances the details in the dark and the bright areas with low computations without boosting noise information and affecting the compressibility of the original image since it performs on the images in the compressed domain. In order to evaluate the proposed scheme, several base-line approaches are described and compared using enhancement quality measures

Properties of the singular value decomposition for efficient data clustering

We introduce some interesting properties of the singular value decomposition (SVD), and illustrate how they may be used in conjunction with the k-means algorithm for efficiently clustering a set of vectors. Specifically, we use the SVD to preprocess and sort the data vectors, and then use the k-means algorithm on the modified vectors. To illustrate the effectiveness of this approach, we compare it to the k-means algorithm without preprocessing and show that significant gains in clustering speed may be realized.

Real-time camera motion classification for content-based indexing and retrieval using templates

In this paper, a simple approach to camera motion analysis is proposed to index videos compressed using MPEG (Moving Picture Experts Group) −1, 2 in faster than real time. Specifically, this paper presents a template-matching algorithm to classify basic camera operations. The proposed approach involves 1) the construction of motion vector fields (MVFs) and filtering out noise; and 2) the sub-MVFs which come from dividing an MVF into several non-overlapping areas for the template matching of camera motions. A fine segmentation also can be obtained for a video, based on the homogeneity of the camera operation in each unit. The advantages of this method lie in real-time processing and robustness to noisy environments such as false motion vectors and object motions. The experimental results show that the proposed algorithm successfully classifies the camera movements.

Content-based image enhancement in the compressed domain based on multi-scale α-rooting algorithm

A simple multi-scale image enhancement algorithm for compressing image dynamics and enhancing image details in the discrete cosine transform (DCT) domain is presented. First, an image is separated into illumination and reflectance components. Next, the illumination component is manipulated adaptively for image dynamics by using a content measure. The content measure using the energy distribution of the DCT coefficients is defined directly in each DCT block of an image. Then, the reflectance component is altered by a multi-scale @a-rooting method for enhancing image details based on human visual perception. The main advantage of the proposed algorithm enhances the details in the dark and the bright areas with low computations without boosting noise information and affecting the compressibility of the original image since it performs on the images in the compressed domain. In order to evaluate the proposed scheme, several base-line approaches are described and compared using enhancement quality measures.

An SSIMD compiler for the implementation of linear shift-invariant flow graphs

This paper describes the implementation of a multiprocessor compiler for signal flow graphs. The compiler takes as input an intrinsically non-parallel representation of a recursive or nonrecursive signal flow graph and generates as output the complete realization for a synchronous multiprocessor computer. Using this compiler, a programmer can generate processor-optimal (optimally efficient) and rate-optimal (optimally fast) implementations without explicitely considering the multiprocessor context or the parallelism.

Fast partial difference elimination algorithm based on block matching error prediction

We introduce a fast partial difference elimination (PDE) algorithm for motion estimation (ME). The basic idea of the proposed approach is to eliminate invalid candidates earlier by predicting a total matching error between matching and candidate blocks. The matching error prediction is performed by using a partially computed matching error between the blocks. Experimental results show that the proposed algorithm can significantly save about 40% of the averaged computational costs versus the conventional PDE algorithm for ME at the cost of ignorable image quality degradation, whose average value is 0.0012 dB.

An application for interactive video abstraction

This paper proposes a novel system that can automatically create an optimal and nonrepetitive summarization and support different user requirements for video browsing and content overview by outputting both the optimal set of key frames and a summarized version of the original video with the user-specified time length. Comparing our approach to video abstraction with another algorithm, we demonstrate that our approach is fast and produces an effective video summary.

A fast clustering algorithm for video abstraction

This paper introduces a useful property of the singular value decomposition (SVD) and uses it to quickly summarize a video sequence based on the visual similarities of its frames. In our method, a video is expressed as the representative frames extracted by a simple key-frame extraction algorithm applied in a sequential manner. Then those key-frames are put together with little redundancy using a clustering algorithm for video abstraction. In order to evaluate the proposed scheme, the speed of the commonly used k-means algorithm for clustering is compared with that of the proposed method that combines both the SVD and the k-means algorithm. Experimental results show that our algorithm is fast and effectively summarizes the content of a video with little redundancy.

A topological sorting and loop cleansing algorithm for a constrained MIMD compiler of shift-invariant flow graphs

This paper describes the use of topological sorting and loop cleansing techniques in an implementation of a multiprocessor compiler for shift invariant flow graphs. The complier is capable of generating three classes of synchronous MIMD implementations: skewed single instruction multiple data (SSIMD), static-parallel skewed single instruction multiple data (Static-PSSIMD), and parallel skewed single instruction multiple data (PSSIMD). The effect of the topological sorting and loop cleansing algorithm is to dramatically reduce the compilation time required to find an optimal or a best SSIMD implementation, as well as maximally fast, perfectly efficient, and minimum delay PSSIMD and Static-PSSIMD implementations.

Dynamic range compression and contrast enhancement for digital images in the compressed domain

We develop a simple and efficient algorithm for dynamic range compression and contrast enhancement of digital images in the compressed domain. The basic idea of our approach is to separate illumination and reflectance components of an image in the compressed domain. We adjust the amount of contribution of the illumination component to effectively compress the dynamic range of the image. For contrast enhancement, we modify the reflectance component based on a new measure of the spectral contents of the image. The spectral content measure is computed from the energy distribution across different spectral bands in a discrete cosine transform (DCT) block. The advantages of the proposed algorithm are (1) high dynamic range scenes are effectively mapped to the smaller dynamic range of the image, (2) the details in very dark or bright areas become clearly visible, (3) the computational cost is low, and (4) the compressibility of the original image is not affected by the algorithm. We evaluate the performance of the proposed algorithm with well-known existing methods, such as histogram equalization and -rooting algorithm, using a few different enhancement quality metrics.