Joon Ki Paik

Recent advances in visual and infrared face recognition: a review

Face recognition is a rapidly growing research area due to increasing demands for security in commercial and law enforcement applications. This paper provides an up-to-date review of research efforts in face recognition techniques based on two-dimensional (2D) images in the visual and infrared (IR) spectra. Face recognition systems based on visual images have reached a significant level of maturity with some practical success. However, the performance of visual face recognition may degrade under poor illumination conditions or for subjects of various skin colors. IR imagery represents a viable alternative to visible imaging in the search for a robust and practical identification system. While visual face recognition systems perform relatively reliably under controlled illumination conditions, thermal IR face recognition systems are advantageous when there is no control over illumination or for detecting disguised faces. Face recognition using 3D images is another active area of face recognition, which provides robust face recognition with changes in pose. Recent research has also demonstrated that the fusion of different imaging modalities and spectral components can improve the overall performance of face recognition.

Contrast enhancement system using spatially adaptive histogram equalization with temporal filtering

In this paper we propose a block-overlapped histogram equalization system for enhancing contrast of image sequences. The proposed system has various applications such as video door phone, security video cameras, in addition to the original target video camcorders.

An adaptive motion decision system for digital image stabilizer based on edge pattern matching

The effects of various environmental conditions which degrade the performance of a digital image stabilization (DIS) system in a video camera are analyzed. On the basis of the analysis, a new DIS system with an adaptive motion decision system is proposed. The DIS system is composed of (i) a local motion vector generation unit, (ii) a field motion vector generation unit, (iii) an accumulated motion vector generation unit, and (iv) field memory address control and a digital zooming unit. >

Normal vector voting: crease detection and curvature estimation on large, noisy meshes

This paper describes a robust method for crease detection and curvature estimation on large, noisy triangle meshes. We assume that these meshes are approximations of piecewise-smooth surfaces derived from range or medical imaging systems and thus may exhibit measurement or even registration noise. The proposed algorithm, which we call normal vector voting, uses an ensemble of triangles in the geodesic neighborhood of a vertex-instead of its simple umbrella neighborhood-to estimate the orientation and curvature of the original surface at that point. With the orientation information, we designate a vertex as either lying on a smooth surface, following a crease discontinuity, or having no preferred orientation. For vertices on a smooth surface, the curvature estimation yields both principal curvatures and principal directions while for vertices on a discontinuity we estimate only the curvature along the crease. The last case for no preferred orientation occurs when three or more surfaces meet to form a corner or when surface noise is too large and sampling density is insufficient to determine orientation accurately. To demonstrate the capabilities of the method, we present results for both synthetic and real data and compare these results to the G. Taubin (1995, in Proceedings of the Fifth International Conference on Computer Vision, pp. 902-907) algorithm. Additionally, we show practical results for several large mesh data sets that are the motivation for this algorithm.

An edge-preserving image interpolation system for a digital camcorder

Conventional image interpolation techniques degrade the quality of the magnified image due to various artifacts, such as, blocking artifact and excessive smoothing. Those degradations become worse as the magnification ratio increases and there also exists a tradeoff between reducing the blocking artifact and excessive smoothness. An image interpolation system, which provides more naturally magnified images than the conventional ones, is proposed. The system preserves the original edge while not destroying the smoothness in the flat area. The hardware structure of the proposed system is also proposed.

Optical flow-based real-time object tracking using non-prior training active feature model

This paper presents a feature-based object tracking algorithm using optical flow under the non-prior training (NPT) active feature model (AFM) framework. The proposed tracking procedure can be divided into three steps: (i) localization of an object-of-interest, (ii) prediction and correction of the objects position by utilizing spatio-temporal information, and (iii) restoration of occlusion using NPT-AFM. The proposed algorithm can track both rigid and deformable objects, and is robust against the objects sudden motion because both a feature point and the corresponding motion direction are tracked at the same time. Tracking performance is not degraded even with complicated background because feature points inside an object are completely separated from background. Finally, the AFM enables stable tracking of occluded objects with maximum 60% occlusion. NPT-AFM, which is one of the major contributions of this paper, removes the off-line, preprocessing step for generating a priori training set. The training set used for model fitting can be updated at each frame to make more robust objects features under occluded situation. The proposed AFM can track deformable, partially occluded objects by using the greatly reduced number of feature points rather than taking entire shapes in the existing shape-based methods. The on-line updating of the training set and reducing the number of feature points can realize a real-time, robust tracking system. Experiments have been performed using several in-house video clips of a static camera including objects such as a robot moving on a floor and people walking both indoor and outdoor. In order to show the performance of the proposed tracking algorithm, some experiments have been performed under noisy and low-contrast environment. For more objective comparison, PETS 2001 and PETS 2002 datasets were also used.

Color active shape models for tracking non-rigid objects

Active shape models can be applied to tracking non-rigid objects in video image sequences. Traditionally these models do not include color information in their formulation. In this paper, we present a hierarchical realization of an enhanced active shape model for color video tracking and we study the performance of both hierarchical and nonhierarchical implementations in the RGB, YUV, and HSI color spaces.

Image interpolation using adaptive fast B-spline filtering

An adaptive version of a B-spline interpolation algorithm is proposed. Adaptivity is used in two different phases: (1) adaptive zero order interpolation is realized by considering directional edge information, and (2) adaptive length of the moving average filter in four directions is obtained by computing the local image statistics. The proposed algorithm exhibits significant improvements in image quality compared with the conventional B-spline type for algorithm, especially with high magnification ratio, such as four times or more. Another advantage of the proposed algorithm is its simplicity in both computation and implementations.<<ETX>>

Robust crease detection and curvature estimation of piecewise smooth surfaces from triangle mesh approximations using normal voting

In this paper, we describe a robust method for the estimation of curvature on a triangle mesh, where this mesh is a discrete approximation of a piecewise smooth surface. The proposed method avoids the computationally expensive process of surface fitting and instead employs normal voting to achieve robust results. This method detects crease discontinuities on the surface to improve estimates near those creases. Using a voting scheme, the algorithm estimates both principal curvatures and principal directions for smooth patches. The entire process requires one user parameter-the voting neighborhood size, which is a function of sampling density, feature size, and measurement noise. We present results for both synthetic and real data and compare these results to an existing algorithm developed by Taubin (1995).

Simple and efficient algorithm for part decomposition of 3-D triangulated models based on curvature analysis

This paper presents a simple and efficient algorithm for part decomposition of compound objects based on Gaussian curvature analysis. The proposed algorithm consists of three major steps, Gaussian curvature estimation, boundary detection, and region growing. Boundaries between two articulated parts are composed of points with highly negative curvature based on the transversality regularity. These boundaries are therefore detected by thresholding estimated Gaussian curvatures for each vertex. A component labeling operation is then performed to grow non-boundary vertices into parts. The original contributions of this paper include: (i) novel, curvature analysis-based decomposition of 3-D models represented by triangle meshes into functional parts instead of surfaces and (ii) large mesh (over 100,000 triangles) handling capability with low computational cost and easy implementation. Experiments were conducted on a large number of both synthetic and real 3-D models. Experimental results demonstrated the performance and efficiency of the proposed algorithm.