• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.9
• /
• pp.1319-1324
• /
• 2013

The mean shift algorithm has achieved considerable success in object tracking due to its simplicity and robustness. It finds local minima of a similarity measure between the color histograms or kernel density estimates of the target and candidate image. However, it is sensitive to the noises due to objects or background having similar color distributions. In addition, occlusion by another object often causes a face region to change in size and position although a face region is a critical clue to perform face recognition or compute face orientation. We assume that depth and color are effective to separate a face from a background and a face from objects, respectively. From the assumption we devised a bilateral filter using color and depth and incorporate it into the mean-shift algorithm. We demonstrated the proposed method by some experiments.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.6
• /
• pp.682-689
• /
• 2019

In part-based image representation, the partial shapes of an object are represented as basis vectors, and an image is decomposed as a linear combination of basis vectors where the coefficients of those basis vectors represent the partial (or local) feature of an object. In this paper, a face recognition for occluded faces is proposed in which face images are represented using non-negative matrix factorization(NMF), one of part-based representation techniques, and recognized using an artificial neural network technique. Standard NMF, projected gradient NMF and orthogonal NMF were used in part-based representation of face images, and their performances were compared. Learning vector quantizer were used in the recognizer where Euclidean distance was used as the distance measure. Experimental results show that proposed recognition is more robust than the conventional face recognition for the occluded faces.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.187-190
• /
• 2022

본 논문은 메타버스 등 인공지능 연계 증강/가상현실 부동 중계 플랫폼에서 부동산 영상 기반 매물 소개 시스템 구축에서 사생활 및 개인정보가 영상에 담기게 될 수 있는 위험이 존재하기에 부동산 영상 내의 개인정보 및 민감 정보를 인공지능 기술을 기반으로 검출하여 삭제해주고 복원해주는 인공지능 기술 연구개발을 목표로 하였다. 한국형 부동산 내 민감 object 를 정의하고, 최신 인공지능 딥러닝 기술 기반 민감 object detection 알고리즘을 연구 개발하며, 영상에서 삭제된 부분은 인공지능 기술을 기반으로 물체가 없는 실제 공간영상으로 복원해주는 영상복원 기술도 연구 개발하였다. 한국형 부동산 환경 (영상 촬영 조도, 디스플레이 스타일, 주변 가구 배치 등)에 맞는 인공지능 모델 구축을 위하여, 자체적으로 한국 영상 database 구축 및 Transfer learning for target domain adaptation 을 진행하였다. 제안된 알고리즘은 일반적인 환경에서 98%의 정확도와 challenge 환경에서 (occlusion 빛 반사, 저조도 등) 81%의 정확도를 보였다. 본 기술은 Proptech 분야에서 주목받고 있는 메타버스 기반 온라인 중계 서비스 기술을 활성화하기 위하여 기획되었으며, 특히 메타버스 부동산 중계 플랫폼의 활성화를 위하여 사생활 보호 측면에서 필요한 중요 기술을 인공지능 기술을 활용하여 연구 개발하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.17-22
• /
• 2009

In this paper, we discuss two fundamental issues of hybrid scene representation: constructing and rendering. A hybrid scene consists of triangular meshes and point-set models. Consider the maturity of modeling techniques of triangular meshes, we suggest that generate a point-set model from a triangular mesh might be an easier and more economical way. We improve stratified sampling by introducing the concept of priority. Our method has the flexibility that one may easily change the importance criteria by substituting priority functions. While many works were devoted to blend rendering results of point and triangle, our work tries to render point-set models and triangular meshes as individuals. We propose a novel way to eliminate depth occlusion artifacts and to texture a point-set model. Finally, we implement our rendering algorithm with the new features of the shader model 4.0 and turns out to be easily integrated with existing rendering techniques for triangular meshes.

• Proceedings of the IEEK Conference
• /
• 2006.06a
• /
• pp.685-686
• /
• 2006

Recently, various AR-based product design methodologies have been introduced. In this paper, we propose technologies for enhancing robust augmentation and immersive realization of virtual objects. A robust augmentation technology is developed for various lighting conditions and a partial solution is proposed for the hand occlusion problem that occurs when the virtual objects overlay the user' hands. It provides more immersive or natural images to the users. Finally, vibratory haptic cues by page motors as well as button clicking force feedback by modulating pneumatic pressures are proposed while interacting with virtual widgets. Also our system reduces gabs between modeling spaces and user spaces. An immersive game-phone model is selected to demonstrate that the users can control the direction of the car in the racing game by tilting a tangible object with the proposed augmented haptic and robust non-occluded visual feedback. The proposed methodologies will be contributed to the immersive realization of the conventional AR system.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.9 no.3
• /
• pp.166-171
• /
• 2009

In recent years, video detection systems have been implemented in various infrastructures such as airport, public transportation, power generation system, water dam and so on. Recognizing moving objects in video sequence is an important problem in computer vision, with applications in several fields, such as video surveillance and target tracking. Segmentation and tracking of multiple vehicles in crowded situations is made difficult by inter-object occlusion. In the system described in this paper, the mean shift algorithm is firstly used to filter and segment a color vehicle image in order to get candidate regions. These candidate regions are then analyzed and classified in order to decide whether a candidate region contains a license plate or not. And then some characters in the license plate is recognized by using the fuzzy ARTMAP neural network, which is a relatively new architecture of the neural network family and has the capability to learn incrementally unlike the conventional BP network. We finally design a license plate recognition system using the mean shift algorithm and fuzzy ARTMAP neural network and show its performance via some computer simulations.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.12C
• /
• pp.1175-1180
• /
• 2005

We present a vision-based hand gesture detection and tracking system. Most conventional hand gesture recognition systems utilize a simpler method for hand detection such as background subtractions with assumed static observation conditions and those methods are not robust against camera motions, illumination changes, and so on. Therefore, we propose a statistical method to recognize and detect hand regions in images using geometrical structures. Also, Our hand tracking system employs multiple cameras to reduce occlusion problems and non-synchronous multiple observations enhance system scalability. In this experiment, the proposed method has recognition rate of $99.28\%$ that shows more improved $3.91\%$ than the conventional appearance method.

• ETRI Journal
• /
• v.37 no.4
• /
• pp.743-751
• /
• 2015

Cloud gaming services are heavily dependent on the efficiency of real-time video streaming technology owing to the limited bandwidths of wire or wireless networks through which consecutive frame images are delivered to gamers. Video compression algorithms typically take advantage of similarities among video frame images or in a single video frame image. This paper presents a method for computing and extracting both graphics information and an object's boundary from consecutive frame images of a game application. The method will allow video compression algorithms to determine the positions and sizes of similar image blocks, which in turn, will help achieve better video compression ratios. The proposed method can be easily implemented using function call interception, a programmable graphics pipeline, and off-screen rendering. It is implemented using the most widely used Direct3D API and applied to a well-known sample application to verify its feasibility and analyze its performance. The proposed method computes various kinds of graphics information with minimal overhead.

• Journal of Electrical Engineering and Technology
• /
• v.9 no.3
• /
• pp.1023-1033
• /
• 2014

In this paper, performance improved hole-filling algorithm which includes the boundary noise removing pre-process that can be used for an arbitrary virtual view synthesis has been proposed. Boundary noise occurs due to the boundary mismatch between depth and texture images during the 3D warping process and it usually causes unusual defects in a generated virtual view. Common-hole is impossible to recover by using only a given original view as a reference and most of the conventional algorithms generate unnatural views that include constrained parts of the texture. To remove the boundary noise, we first find occlusion regions and expand these regions to the common-hole region in the synthesized view. Then, we fill the common-hole using the spiral weighted average algorithm and the gradient searching algorithm. The spiral weighted average algorithm keeps the boundary of each object well by using depth information and the gradient searching algorithm preserves the details. We tried to combine strong points of both the spiral weighted average algorithm and the gradient searching algorithm. We also tried to reduce the flickering defect that exists around the filled common-hole region by using a probability mask. The experimental results show that the proposed algorithm performs much better than the conventional algorithms.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.5
• /
• pp.389-395
• /
• 2019

Unlike aerial images covering small region, satellite data show high potential to detect urban scale geospatial changes. The change detection using satellite images can be carried out using single image or stereo images. The single image approach is based on radiometric differences between two images of different times. It has limitations to detect building level changes when the significant occlusion and relief displacement appear in the images. In contrast, stereo satellite data can be used to generate DSM (Digital Surface Model) that contain information of relief-corrected objects. Therefore, they have high potential for the object change detection. Therefore, we carried out a study for the change detection over an urban area using stereo satellite data of two different times. First, the RPC correction was performed for two DSMs generation via stereo image matching. Then, D-DSM (Differential DSM) was generated by differentiating two DSMs. The D-DSM was used for the topographic change detection and the performance was checked by applying different height thresholds to D-DSM.