• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.4
• /
• pp.2075-2092
• /
• 2017

Person re-identification is to match pedestrians observed from non-overlapping camera views. It has important applications in video surveillance such as person retrieval, person tracking, and activity analysis. However, it is a very challenging problem due to illumination, pose and viewpoint variations between non-overlapping camera views. In this work, we propose a viewpoint invariant method for matching pedestrian images using orientation of pedestrian. First, the proposed method divides a pedestrian image into patches and assigns angle to a patch using the orientation of the pedestrian under the assumption that a person body has the cylindrical shape. The difference between angles are then used to compute the similarity between patches. We applied the proposed method to real-time global multi-object tracking across multiple disjoint cameras with non-overlapping field of views. Re-identification algorithm makes global trajectories by connecting local trajectories obtained by different local trackers. The effectiveness of the viewpoint invariant method for person re-identification was validated on the VIPeR dataset. In addition, we demonstrated the effectiveness of the proposed approach for the inter-camera multiple object tracking on the MCT dataset with ground truth data for local tracking.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.8
• /
• pp.1433-1445
• /
• 1994

In this paper, an coherent optical correlator system based on the off-axis projection synthetic discriminant funtion (OA-pSDF) was analyzed and implemented optically. The filter was synthesized by combining conventional pSDF with single reference plane wave multiplexing. Synthesized pSDF were transformed to binary phase only filters (BPOFs) and fabricated as computer generated holograms(CGHs), which was used in the real time optical correlator system instead of using expensive spatial light modulators(SLMs). From the characteristic test, it was found that OA-pSDF showed distortion invariance and good performances in discrminating subset images. The proposed OA-pSDF BPOF could overcome the limitations of conventional BPOFs : that is distortion variance such as acale and rotation, especially out of plane variance.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.13 no.2
• /
• pp.213-220
• /
• 1989

This paper describes the application of the variable structure control(VSC) concept for the position control of an electro-hydraulic servomotor system. The basic philosopy of VSC is that the structure of the feedback control is altered as the state crosses discontinuity surfaces in the state space with the result that certain desirable properties are achieved. The switching of the control function yields total(or selective) invariance to system parameter variations and disturbances, and closed loop eigen value placement in time-varing and uncertain systems. The control scheme is derived, implemented and tested in the laboratory where analog controller have been used to control the representive servosystem. The control system schematics are given and simple results are shown for illustration. And the results of variable structure system for the electro-hydraulic servomotor were compared to that of the fixed structure system when load disturbance and system parameter variation exists.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.9 no.1
• /
• pp.165-176
• /
• 2013

This paper introduces the high density discrete wavelet transform using quincunx sampling, which is a discrete wavelet transformation that combines the high density discrete transformation and non-separable processing method, each of which has its own characteristics and advantages. The high density discrete wavelet transformation is one that expands an N point signal to M transform coefficients with M > N. The high density discrete wavelet transformation is a new set of dyadic wavelet transformation with two generators. The construction provides a higher sampling in both time and frequency. This new transform is approximately shift-invariant and has intermediate scales. In two dimensions, this transform outperforms the standard discrete wavelet transformation in terms of shift-invariant. Although the transformation utilizes more wavelets, sampling rates are high costs and some lack a dominant spatial orientation, which prevents them from being able to isolate those directions. A solution to this problem is a non separable method. The quincunx lattice is a non-separable sampling method in image processing. It treats the different directions more homogeneously than the separable two dimensional schemes. Proposed wavelet transformation can generate sub-images of multiple degrees rotated versions. Therefore, This method services good performance in image processing fields.

• Proceedings of the IEEK Conference
• /
• 2002.07c
• /
• pp.1578-1581
• /
• 2002

In this paper, the accuracy of the direction-of-arrival (DOA) estimation of signal impinged on the uniform linear array (ULA) is investigated. The conventional beamformer and Capon’s beamformer categorized in beamformaing techniques as well as MUSIC (MUlti-pie Signal Classification) and ESPRIT (Estimation of Signal Invariance Techniques) categorized in subspace- based methods are employed to estimate the DOAs. From the simulation result under uncorrelated environment, MUSIC can prominently distinguish the DOAs while the beamforming techniques cannot demonstrate the DOAs as clear as MUSIC does. Moreover, Uni-tary ESPRIT is employed to estimate the DOAs under uncorrelated signal conditions. By means of Uni-tary ESPRIT, the estimation has more accuracy with the computational-time reduction. In addition, it incorporates forward-backward averaging; thus Unitary ES-PRIT can overcome the problem of the coherent signal condition.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.8 no.3
• /
• pp.133-145
• /
• 2012

This paper describes the high density discrete wavelet transformation which is one that expands an N point signal to M transform coefficients with M > N. The double-density discrete wavelet transform is one of the high density discrete wavelet transformation. This transformation employs one scaling function and two distinct wavelets, which are designed to be offset from one another by one half. And it is nearly shift-invariant. Similarly, triple-density discrete wavelet transformation is a new set of dyadic wavelet transformation with two generators. The construction provides a higher sampling in both time and frequency. Specifically, the spectrum of the first wavelet is concentrated halfway between the spectrum of the second wavelet and the spectrum of its dilated version. In addition, the second wavelet is translated by half-integers rather than whole-integers in the frame construction. This arrangement leads to high density wavelet transformation. But this new transform is approximately shift-invariant and has intermediate scales. In two dimensions, this transform outperforms the standard and double-density discrete wavelet transformation in terms of multiple directions. Resultingly, the proposed wavelet transformation services good performance in image and video processing fields.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.4
• /
• pp.2094-2112
• /
• 2019

Focusing on the issue that the conventional Kernel Correlation Filter (KCF) algorithm has poor performance in handling scale change and obscured objects, this paper proposes an anti-occlusion and scale adaptive tracking algorithm in the basis of KCF. The average Peak-to Correlation Energy and the peak value of correlation filtering response are used as the confidence indexes to determine whether the target is obscured. In the case of non-occlusion, we modify the searching scheme of the KCF. Instead of searching for a target with a fixed sample size, we search for the target area with multiple scales and then resize it into the sample size to compare with the learnt model. The scale factor with the maximum filter response is the best target scaling and is updated as the optimal scale for the following tracking. Once occlusion is detected, the model updating and scale updating are stopped. Experiments have been conducted on the OTB benchmark video sequences for compassion with other state-of-the-art tracking methods. The results demonstrate the proposed method can effectively improve the tracking success rate and the accuracy in the cases of scale change and occlusion, and meanwhile ensure a real-time performance.

• The KIPS Transactions:PartA
• /
• v.10A no.4
• /
• pp.357-368
• /
• 2003

Evaluation of shape similarity for 3D models is essential in many areas - medicine, mechanical engineering, molecular biology, etc. Moreover, as 3D models are commonly used on the Web, many researches have been made on the classification and retrieval of 3D models. In this paper, we describe methods for 3D shape representation and major concepts of similarity evaluation, and analyze the key features of recent researches for shape comparison after classifying them into four categories including multi-resolution, topology, 2D image, and statistics based methods. In addition, we evaluated the performance of the reviewed methods by the selected criteria such as uniqueness, robustness, invariance, multi-resolution, efficiency, and comparison scope. Multi-resolution based methods have resulted in decreased computation time for comparison and increased preprocessing time. The methods using geometric and topological information were able to compare more various types of models and were robust to partial shape comparison. 2D image based methods incurred overheads in time and space complexity. Statistics based methods allowed for shape comparison without pose-normalization and showed robustness against affine transformations and noise.

• Journal of Internet Computing and Services
• /
• v.6 no.6
• /
• pp.187-197
• /
• 2005

In this paper, we propose a method to estimate camera motion parameter based on invariant point features, Typically, feature information of image has drawbacks, it is variable to camera viewpoint, and therefore information quantity increases after time, The LM(Levenberg-Marquardt) method using nonlinear minimum square evaluation for camera extrinsic parameter estimation also has a weak point, which has different iteration number for approaching the minimal point according to the initial values and convergence time increases if the process run into a local minimum, In order to complement these shortfalls, we, first proposed constructing feature models using invariant vector of geometry, Secondly, we proposed a two-stage calculation method to improve accuracy and convergence by using 2D homography and LM method, In the experiment, we compared and analyzed the proposed method with existing method to demonstrate the superiority of the proposed algorithms.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.783-788
• /
• 2003

In this paper, we present a new approach to detect and recognize human face in the image from vision camera equipped on the mobile robot platform. Due to the mobility of camera platform, obtained facial image is small and pose-various. For this condition, new algorithm should cope with these constraints and can detect and recognize face in nearly real time. In detection step, ‘coarse to fine’ detection strategy is used. Firstly, region boundary including face is roughly located by dual ellipse templates of facial color and on this region, the locations of three main facial features- two eyes and mouth-are estimated. For this, simplified facial feature maps using characteristic chrominance are made out and candidate pixels are segmented as eye or mouth pixels group. These candidate facial features are verified whether the length and orientation of feature pairs are suitable for face geometry. In recognition step, pseudo-convex hull area of gray face image is defined which area includes feature triangle connecting two eyes and mouth. And random lattice line set are composed and laid on this convex hull area, and then 2D appearance of this area is represented. From these procedures, facial information of detected face is obtained and face DB images are similarly processed for each person class. Based on facial information of these areas, distance measure of match of lattice lines is calculated and face image is recognized using this measure as a classifier. This proposed detection and recognition algorithms overcome the constraints of previous approach [15], make real-time face detection and recognition possible, and guarantee the correct recognition irregardless of some pose variation of face. The usefulness at mobile robot application is demonstrated.