Seong Dae Kim

A bit allocation method based on picture activity for still image coding

Bit allocation or the quantizer assignment problem is a basic and essential issue in lossy picture coding. The optimal solution for the bit allocation problem can be found by the Lagrangian method. However, it requires much computational time and memory. To reduce complexity overhead, we propose a fast scheme using a picture activity measure. Comparison among the existing activity measurement methods is presented to select the most reliable activity measure and the mapping relation between the activity value, and a quantization parameter is proposed.

Scale-invariant template matching using histogram of dominant gradients

Abstract This paper presents a histogram-based template matching method that copes with the large scale difference between target and template images. Most of the previous template matching methods are sensitive to the scale difference between target and template images because the features extracted from the images are changed according to the scale of the images. To overcome this limitation, we introduce the concept of dominant gradients and describe an image as the feature that is tolerant to scale changes. To this end, we first extract the dominant gradients of a template image and represent the template image as the grids of histograms of the dominant gradients. Then, the arbitrary regions of a target image with various locations and scales are matched with the template image via histogram matching. Experimental results show that the proposed method is more robust to scale difference than previous template matching techniques.

An efficient real time focusing delay calculation in ultrasonic imaging systems

An efficient real time focusing delay calculation algorithm is proposed for variable sampling clock generation (SCG) with high accuracy needed in digital focusing in ultrasonic imaging systems. The proposed algorithm is an extension of the midpoint drawing algorithm that is well known in the computer graphics area. It can be implemented with simple hardware amenable to VLSI realization, without using a large amount of look-up memory to store the sampling clock information otherwise required.

Dynamic background subtraction via sparse representation of dynamic textures in a low-dimensional subspace

In this paper, we deal with the problem of background subtraction especially for the scenes containing dynamic textures. In the scenes, unlike static textures, dynamic textures show a wide range of per-pixel color variations over time. For successful dynamic background subtraction, therefore, it is an essential task to represent the dynamics of these variations effectively. For this task, in the proposed method, i) a training set of dynamic background scenes is modeled in a low-dimensional subspace and then ii) the background of a test scene is represented as a linear combination of a few coefficient matrices resulting from the projection of the training scenes onto the low-dimensional subspace. More specifically, the proposed dynamic background subtraction method is based on the sparse representation of dynamic textures in the low-dimensional subspace. In the experiments, the proposed method shows promising performance in comparison with other competitive methods in the literature.

Abrupt shot change detection using an unsupervised clustering of multiple features

We propose an efficient method to detect abrupt shot changes in a video sequence by using an unsupervised clustering. Most conventional shot change detection algorithms use only one kind of frame-by-frame difference feature such as pixel difference or histogram difference, so they can be applied to only specific situations. Another problem is the determination of appropriate threshold values to check the existence of shot changes. To overcome these problems we use several kinds of features simultaneously and propose a modified k-means clustering algorithm which changes the initial cluster center adaptively. Experimental results show that the proposed algorithm works well.

A fast and adaptive method to estimate texture statistics by the spatial gray level dependence matrix (SGLDM) for texture image segmentation

Abstract This paper presents a fast and adaptive algorithm to estimate texture statistics based on the SGLDM for texture image segmentation. We propose several recursive equations by modifying the conventional ones in order to get the texture statistics at high speed. And also to find the more accurate texture statistics along the boundaries between regions, a locally adaptive method to obtaining a suitable size of a window is described. To evaluate the performance of the proposed method, it is simulated using an artificial image. By the experimental results, we conclude that the proposed method is useful to estimate texture statistics for image segmentation.

Recursive On-Line

In this paper, we propose a novel background subtraction method which enables reliable detection of foreground objects in a long surveillance video stream. Recently, although much progress has been made in the field of background subtraction, there are still challenging scenarios (e.g., high frequency motion of dynamic texture, non-stationary motion of camera, abrupt changes of illumination, etc.) in the long surveillance videos in which even state-of-the-art methods are often prone to fail. To cope with these challenging scenarios effectively, in the proposed method, a background model is initialized in a low-dimensional subspace and then updated periodically based on a novel recursive on-line (2D)2PCA algorithm developed in this paper. Moreover, a threshold map is also updated in a scene-adaptive manner for labeling each pixel in a scene either foreground or background independently. Based on this on-line framework, the background of a surveillance video stream is reconstructed over time, thereby facilitating the detection of foreground objects reliably. In extensive experiments, we demonstrate that the proposed background subtraction method can cope with the aforementioned challenging scenarios more favorably than the state-of-the-art methods.

{(2{\rm D})}^2{\rm PCA}

In this paper, we present a novel color-to-gray image conversion method which preserves both color and texture discriminabilities effectively. Unlike previous approaches, the proposed method does not require any user-specific parameters for conversion. Moreover, the computational complexity is low enough to be applied to real-time applications. These breakthroughs are achieved by applying the ELSSP (Eigenvalue-weighted Linear Sum of Subspace Projections) method, which is proposed in this paper for the color-to-gray image conversion. Experimental results demonstrate that the proposed method is superior to the state-of-the-art methods in terms of both conversion speed and image quality.

and Its Application to Long-Term Background Subtraction

We propose new face recognition algorithm using 3D line edge map as combining 3D information on the edge obtained by stereo camera and the conventional line edge map algorithm. As a similarity measure of 3D line edge map, weighted partial and spatially well-localized line segment Hausdorff distance (WPSLHD) is proposed. WPSLHD is a measure which is robust to expression changes, facial occlusions, and depth outliers. The proposed face recognition system performances are evaluated under expression changes and facial occlusions.

Novel PCA-based color-to-gray image conversion

As image-based 3-D modeling is used in a variety of applications, accordingly, the compression of 3-D object geometry represented by multiple images becomes an important task. This paper presents a model-based approach to predict the geometric structure of an object using its visual hull. A visual hull is a geometric entity generated by shape-from-silhouette (SFS), and consequently it largely follows the overall shape of the object. The construction of a visual hull is computationally inexpensive and a visual hull can be encoded with relatively small amount of bits because it can be represented with 2-D silhouette images. Therefore, when it comes to the predictive compression of objects geometric data, the visual hull should be an effective predictor. In the proposed method, the geometric structure of an object is represented by a layered depth image (LDI), and a visual hull from the LDI data is computed via silhouette generation and SFS. The geometry of an object is predicted with the computed visual hull, and the visual hull data with its prediction errors are encoded. Simulation results show that the proposed predictive coding based on visual hull outperforms the previous image-based methods and the partial surface-based method.