Yang-ck Seo

Binary shape coding using baseline-based method

Here, we propose a new shape-coding algorithm called baseline-based binary shape coding (BBSC), where the outer or inner contours of an arbitrarily shaped object are represented by traced one-dimensional (l-D) data from the baseline with turning points (TPs). There are two coding modes, i.e., the intra and inter modes as in texture coding. In the intra mode, the differential values of the neighboring 1-D distance values and TPs corresponding to the given shape are encoded by the entropy coder. In the inter mode, object identification, global shape matching and local contour matching are employed for motion compensation/estimation. Lossy shape coding is enabled by variable sampling in each contour segment or by allowing some predefined error when performing motion compensation. We compare the proposed method with the bitmap-based method of context-based arithmetic encoding (CAE). Simulation results show that the proposed method is better than CAE in coding efficiency for intra mode and better in subjective quality for both intra and inter modes, although the CAE method has performed better than the proposed method in inter mode.

Multi-channel audio reproduction apparatus and method for loudspeaker sound reproduction using position adjustable virtual sound images

A multi-channel audio reproduction apparatus and method for loudspeaker reproduction using virtual sound images whose positions can be adjusted is provided. The multi-channel audio reproduction apparatus includes a virtual sound image forming unit for compensating for the occurrence of cross-talk in at least one input audio signal according to the arrangement of loudspeakers, obtaining transfer functions occurring when sound from a position in a three dimensional space is transmitted to both ears of a listener, and forming a plurality of first virtual sound images in a three dimensional space using the transfer functions. A controller generates adjusting factors for adjusting the position of at least one second virtual sound image. An output position adjustor controls the at least one audio signal, with respect to which the plurality of first virtual sound images are formed by the virtual sound image forming unit, with the adjusting factors generated by the controller and adjusts positions of the at least one second virtual sound image. An adder sums up left output related signals of the at least one audio signal with respect to which the position of the at least one second virtual sound image is adjusted, and sums up right output related signals of the at least one audio signal with respect to which the position of the at least one second virtual sound image is adjusted, to generate left and right audio signals for forming the at least one second virtual sound image.

An effective news anchorperson shot detection method based on adaptive audio/visual model generation

A multi-modal method to improve the performance of the anchorperson shot detection for news story segmentation is proposed in this paper. The anchorperson voice information is used for the verification of anchorperson shot candidates extracted by visual information. The algorithm starts with the anchorperson voice shot candidate extraction using time and silence condition. The anchorperson templates are generated from the anchorperson face and cloth information from the anchorperson voice shots extracted. The anchorperson voice models are then created after segregating anchorperson voice shots containing 2 or more voices. The anchorperson voice model verifies the anchorperson shot candidates obtained from visual information. 720 minutes of news programs are tested and experimental results are demonstrated.

Linear model of surface and scanner characterization method

In color reproduction research, a linear model designed to minimize the error between original surface reflectance spectra and reproduced spectra is useful in the process of producing an accurate color match between the original image and reproduction under a variety of illuminants, but it is inappropriate in efficiency. We propose an efficient linear model based on surface reflectance spectra and a unified wavelength function of CIE 1931 standard observer representing human perceptual property. The surface spectra weighted with the unified wavelength function were introduced to minimize the human perceptual error between original reflectance spectra and reproduced spectra and to reduce the number of the spectral basis functions. The performance of reflectance spectra-to-CIELAB transformation on our proposed linear model is tested and compared with a conventional model based on reflectance spectra under a variety of illuminants. The results of our linear model is superior to that of the conventional model. With Munsell 400 color patches, D65 illuminant and 4-dimensional linear model, the mean color difference of our model is 1.28 CIELAB unit. And an algorithm for color scanner characterization using our model is made and tested, and the results are shown.

Rate control strategy based on human visual sensitivity for MPEG video coder

The Moving Picture Experts Group (MPEG) has standardized the bit-stream syntax for the coded representation of video. This means that MPEG specifies only a decoding method and allows much flexibility in encoding methods. Therefore the picture quality of the reconstructed video sequences is considerably dependent on the rate control strategy in the encoding method. We propose a new rate control strategy conforming to the MPEG syntax that improves the reconstructed picture quality. The new rate control strategy allocates the target number of bits and assigns quantization step sizes adaptively based on the human visual sensitivity and the complexity of the picture to be coded. The proposed rate control strategy consists of the following steps. First, a 16 X 16 macroblock is classified into one of 8 macroblock classes based on human visual sensitivity for luminance and color components in the macroblock and then an 8 X 8 block is classified into one of block classes by its variance. Next, the target number of bits is allocated to a block and the quantization step size is assigned to a macroblock. The result of subjective tests showed the proposed rate control strategy improved the picture quality of the reconstructed video sequences considerably over the conventional strategy. Especially in the proposed strategy, the subjective picture quality between frames was more constant and hence was less degraded than conventional rate control strategies.

Scan interleaving based scalable binary shape coding

Abstract In this paper, we propose scan interleaving based shape coding (SISC) as a spatial scalable coding algorithm for binary shape. A typical scalable video system has two or more scalable layers. These are the base layer and the associated enhancement layers. In the proposed spatial scalable binary shape coding algorithm, scan interleaving based data encoding (SI) scheme is proposed for the enhancement layer coding. And context-based arithmetic encoding (CAE) method is employed for entropy coding of binary shape data in each scalable layer. SI scheme efficiently exploits the spatial redundancy of the scalable layers. For reducing the temporal redundancy, inter-frame coding based on motion compensation is employed. The major advantages of SISC are high coding gain, low complexity, and consistency with MPEG-4 non-scalable binary shape coding. High coding gain is achieved from the method of coded-scan-line data analysis, which uses exclusive-OR operation on top and bottom reference pixels. Furthermore, by referring the lower layer motion vectors for its inter-frame coding process, SISC also saves some computational overhead of enhancement layer coding. The proposed technique is, currently, adopted in the Working Draft of MPEG-4 Version 2. SISC supports both the frame-based coding mode and the block-based coding mode. The block-based SISC, as in MPEG-4, is mainly described in this paper.

New algorithm for detecting illuminant chromaticity from color images

In this paper, the method to calculate the illuminant chromaticity of an image is proposed by combine the perceived illumination and highlight approach. The hybrid approach is more stable and accurate compared to each approach. The application for this algorithm is two-fold. For simple and quick implementation, perceived illumination is enough, and for more accurate case, hybrid approach can be used. And conversion of image illuminant chromaticity is also proposed. This can be applied into special effect for the images.

New image segmentation method using mode finding, multi-link clustering, and region graph analysis

A new approach to image segmentation is presented. Novelty consists in combining multiple image feature information together -- color feature, texture feature and pixel’s geometric location in spatial domain to separate the regions with homogeneous color, texture, and similar spatiality --, as well as grouping the homogeneous clusters in the feature space with unique manner. The proposed segmentation algorithm contains two main stages. First, the mode finding and multi-link clustering algorithm converts an image into a map of small primary regions - region graph representation. The nodes of the graph correspond to distinguished regions, and the lines correspond to relations between neighbor regions. The region map is further simplified by the secondary graph analysis and merging of neighbor regions. The performance of developed algorithm was tested by using various images obtained by a real camera.