• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2010.07a
• /
• pp.70-73
• /
• 2010

최근 IT 기술이 융합된 지능형 자동차 기술에 대한 관심이 높아짐에 따라 이에 대한 연구가 활발히 진행되고 있다. 차선 검출은 지능형 자동차의 주요 과제인 첨단 안전자동차 기술의 핵심적인 부분으로 국내외에서 다양한 방법들에 대한 연구가 진행되었다. 차량의 안전을 향상시키기 위해서는 충분한 제동거리 확보가 가능한 거리까지 정확하고 빠른 차선 검출이 이루어져야 한다. 기존의 경계선 검출기반 차선 검출은 소실점 근처에서 경계선 검출이 이루어지지 않았다. 이는 차선과 도로의 색이 잘 구분되지 않는 채널을 사용하는 문제에서 기인한다. 따라서 본 논문에서는 선형 판별 분석 기법을 이용하여 차선과 도로 색을 가장 잘 구분할 수 있는 RGB 가중치를 계산하여 이로부터 projection 영상을 만들고, 변환한 영상에서 경계선 검출을 수행함으로써 보다 정확한 경계선 검출 결과를 얻는 방법을 제안한다. 제안한 방법으로 얻은 영상과 기존의 흑백 영상에 동일한 경계선 검출기를 적용하여 성능을 비교하고, 이를 적용한 차선검출 실험결과를 제시한다.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.3
• /
• pp.265-274
• /
• 2001

In this paper, we propose a content-based face retrieval system which can retrieve a face based on a facial feature region. Instead of using keyword such as a resident registration number or name for a query, the our system uses a facial image as a visual query. That is, we recognize a face based on a specific feature region including eyes, nose, and mouth. For this, we extract the feature region using the color information based on HSI color model and the edge information from wavelet transformed image, and then recognize the feature region using neural network. The proposed system is implemented on client/server environment based on Oracle DBMS for a large facial image database. In the experiment with 150 various facial images, the proposed method showed about 88.3% recognition rate.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.169-172
• /
• 2009

In this paper, we propose the method to improving touch recognition using edge mask on a touched object at vision-based touchscreen. Because vision-based touchscreen recognizes touch using threshold simply, noise occur in fist or wrist in case of touch directly with hand, correct touch recognition was difficult. However, in this paper, we execute morphology and extract surrounding mask in object that approximate to touchscreen, use change of contrast for the mask. When we touch screen to use these dynamic information, prevent noise. The goal of this paper is when hand was touched on screen it can recognize to touch.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.2B
• /
• pp.352-360
• /
• 2010

Supporting user perceptual QoE-guaranteed IP-based multimedia service such as IPTV and Mobile IPTV, we represent an efficient QoE-aware transport stream assessment schemes to apply realtime mobile IPTV's contents distorted by various network errors such as bandwidth, delay, jitter, and packet loss. This paper proposes in detail an efficient matching and QoE-aware measurement methods. The brightness of the digitized contents per each frames of transport streams is used and applied to reduced-reference method. The hybrid video quality metric is designed by QoE-indicators such as blur, block, edge busyness, and color error. We compare original with processed source to evaluate them in a high precision degree of accuracy.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.1
• /
• pp.236-248
• /
• 1998

Human observers perceive a contour where no line or edge is given, such as the border between gratings of thin lines offset by half a cycle. Illusory contours occur in a wide variety of circumstances in nature. A common factor in all such figures is the perception of a surface or contour occluding part of a background, i.e. illusory contours are always accompanied by illusory surfaces. Some examples about these kinds of figures are given by Kanizsa's triangle or square. In this paper, we proposed a neural network model for forming(extracting) the illusory surfaces, based on the mechanism of feature extraction found in a mammals' visual pathway. It will be shown that the model can extract the illusory surfaces from illusory contours successfully.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.8
• /
• pp.818-823
• /
• 2008

In this paper, we propose a feature extraction method which extracts directional features of handwritten numerals by using the projection runlength. Our directional featrures are obtained from four directional images, each of which contains horizontal, vertical, right-diagonal and left-diagonal lines in entire numeral shape respectively. A conventional method which extracts directional features by using Kirsch masks generates edge-shaped double line directional images for four directions, whereas our method uses the projections and their runlengths for four directions to produces single line directional images for four directions. To obtain the directional projections for four directions from a numeral image, some preprocessing steps such as thinning and dilation are required, but the shapes of resultant directional lines are more similar to the numeral lines of input numerals. Four [$4{\times}4$] directional features of a numeral are obtained from four directional line images through a zoning method. By using a hybrid feature which is made by combining our feature with the conventional features of a mesh features, a kirsch directional feature and a concavity feature, higher recognition rates of the handwrittern numerals can be obtained. For recognition test with given features, we use a multi-layer perceptron neural network classifier which is trained with the back propagation algorithm. Through the experiments with the handwritten numeral database of Concordia University, we have achieved a recognition rate of 97.85%.

• Journal of Digital Convergence
• /
• v.12 no.6
• /
• pp.237-243
• /
• 2014

In this paper, we propose a mobile image retrieval system which utilizes the mobile device's sensor information and enables running in a variety of the environments, and implement the system on Android platform. The proposed system deals with a new image descriptor using combination of the visual feature with EXIF attributes in the target of JPEG image, and image matching algorithm which is optimized to the mobile environments. Experiments are performed on the Android platform, and the experimental results revealed that the proposed algorithm exhibits a significant improved results with large image database.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.9
• /
• pp.95-103
• /
• 1998

This paper proposes an image segmentation algorithm based on fusion of range and intensity images. Based on Bayesian theory, a priori knowledge is encoded by the Markov random field (MRF). A maximum a posteriori (MAP) estimator is constructed using the features extracted from range and intensity images. Objects are approximated by local planar surfaces in range images, and the parametric space is constructed with the surface parameters estimated pixelwise. In intensity images the ${\alpha}$-trimmed variance constructs the intensity feature. An image is segmented by optimizing the MAP estimator that is constructed using a likelihood function based on edge information. Computer simulation results shw that the proposed fusion algorithm effectively segments the images independentl of shadow, noise, and light-blurring.

• Journal of Information Technology Application
• /
• v.3 no.4
• /
• pp.119-134
• /
• 2001

Improving software product quality is a key to increasing user satisfaction and to achieving competitive edge. There are two approaches to assure high software product quality; development process-oriented and product-oriented. There have been many efforts for improving software quality through process certification, for example, CMM, ISO 9000 family, ISO/IEC 12207, SPICE and Bootstrap. However, a good process alone cannot guarantee good product quality. A need for the evaluation of software product quality by an independent third party is growing rapidly for several reasons. We are concerned with an application of Statistical Quality Control (SQC) to the evaluation of software product quality to obtain the efficiency of evaluation processes and the objectivity of evaluation results. Methods for selecting test cases using a random sampling approach have been discussed and methods for selecting acceptance criteria with respect to software product quality have also been suggested.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.36C no.11
• /
• pp.46-54
• /
• 1999

The output values of cellular neural networks would have errors because they can be stabilized at local minimums depending on the initial states of each cell. So, in this paper, we design the $6\times6$cellular neural networks with annealing capability which guarantees that the outputs reaches the global minimum to have correct output values independent of the initial states of each cell. This chip is designed using a $0.8\mu\textrm{m}$ CMOS technology The designed chip contains about 15,000 transistors and the chip size is about $2.89\times2.89\textrm{mm}^2$. The simulation results of edge extraction and hole filling using the designed circuit show that the outputs values would have errors in un-annealed case, but not in annealed case. In the simulation, the annealing time of $3\musec$ is employed.