• 제어로봇시스템학회논문지
• /
• 제7권6호
• /
• pp.523-531
• /
• 2001

This paper describes and automatic mirror adjustment system that rotates a pair of side mirrors and the room mirror of a car to the optimal position for a driver by using the locating of the driver\`s pupils. A stereo vision system measures 3D coordinates of a pair pupils by analyzing the input images of stereo B/W CCD cameras mounted on the instrument panel. this system determines the position angle of each mir-ror on the basis of information about the location of the pupils and rotates each mirror to the appropriate po-sition by mirror actuators. The vision system can detect the driver\`s pupils regardless of whether it is day-time or nighttime by virtue of an infrared light source. information about the pair of nostrils in used to im- prove the correctness of pupil detection. This system can adjust side mirrors and the room mirror automati- cally and rapidly by a simple interface regardless of driver replacement of driver\`s posture. Experiment has shown this to be a new mirror adjustment system that can make up for the weak points of previous mirror adjustment systems.

• 전자공학회논문지
• /
• 제54권5호
• /
• pp.65-73
• /
• 2017

본 논문에서는 홀 영역과 시차 불연속 영역을 개선하기 위한 스테레오 정합 기법을 제안한다. 스테레오 정합 기법은 두 영상에서의 정합 점을 탐색하여 시차 지도를 추출한다. 하지만 기존의 스테레오 정합 기법들은 스테레오 영상의 베이스 라인 길이에 따라서 정확도와 정밀도가 반비례하는 문제점이 있다. 또한 영상의 폐색 영역과 특징 부족으로 인한 시차 불연속 영역이 존재한다. 제안한 기법에서는 개선된 AD-Census-Gradient 방법과 적응적 가중치 기반의 비용 결합을 통하여 불연속 영역과 오 정합 영역을 개선한 초기 시차 지도를 추출하였다. 그 후에 시차 지도 재생성 과정을 수행하여 오정합 영역을 개선함과 동시에 영상의 정밀도를 개선하였다. 실험 결과 제안하는 기법이 기존의 정합률이 높은 방법들과 비교하여 높은 수준의 정합률을 유지하면서 오정합 영역과 정밀도를 개선하였음을 보였다. 그리고 정합 오차율이 높은 영상의 경우, 최근에 발표된 스테레오 정합 방법들보다 정합 성능이 평균적으로 3.22(%)가량 증가하였다.

• Journal of Information Processing Systems
• /
• 제13권2호
• /
• pp.256-267
• /
• 2017

An active research area in computer vision, stereo matching is aimed at obtaining three-dimensional (3D) information from a stereo image pair captured by a stereo camera. To extract accurate 3D information, a number of studies have examined stereo matching algorithms that employ adaptive support weight. Among them, the adaptive census transform (ACT) algorithm has yielded a relatively strong matching capability. The drawbacks of the ACT, however, are that it produces low matching accuracy at the border of an object and is vulnerable to noise. To mitigate these drawbacks, this paper proposes and analyzes the features of an improved stereo matching algorithm that not only enhances matching accuracy but also is also robust to noise. The proposed algorithm, based on the ACT, adopts the truncated absolute difference and the multiple sparse windows method. The experimental results show that compared to the ACT, the proposed algorithm reduces the average error rate of depth maps on Middlebury dataset images by as much as 2% and that is has a strong robustness to noise.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2005년도 Proceedings of ISRS 2005
• /
• pp.667-670
• /
• 2005

Ground control points(GCPs) can be extracted from SAR data given precise orbit for DTM generation using optic images and other SAR data. In this study, we extract GCPs from ERS SAR data and SRTM DEM. Although it is very difficult to identify GCPs in ERS SAR image, the geometry of optic image and other SAR data are able to be corrected and more precise DTM can be constructed from stereo optic images. Twenty GCPs were obtained from the ERS SAR data with precise Delft orbit information. After the correction was applied, the mean values of planimetric distance errors of the GCPs were 3.7m, 12.1 and -0.8m with standard deviations of 19.9m, 18.1, and 7.8m in geocentric X, Y, and Z coordinates, respectively. The geometries of SPOT stereo pair were corrected by 13 GCPs, and r.m.s. errors were 405m, 705m and 8.6m in northing, easting and height direction, respectively. And the geometries of RADARS AT stereo pair were corrected by 12 GCPs, and r.m.s. errors were 804m, 7.9m and 6.9m in northing, easting and height direction, respectively. DTMs, through a method of area based matching with pyramid images, were generated by SPOT stereo images and RADARS AT stereo images. Comparison between points of the obtained DTMs and points estimated from a national 1 :5,000 digital map was performed. For DTM by SPOT stereo images, the mean values of distance errors in northing, easting and height direction were respectively -7.6m, 9.6m and -3.1m with standard deviations of 9.1m, 12.0m and 9.1m. For DTM by RADARSAT stereo images, the mean values of distance errors in northing, easting and height direction were respectively -7.6m, 9.6m and -3.1m with standard deviations of 9.1m, 12.0m and 9.1m. These results met the accuracy of DTED level 2

• 대한원격탐사학회지
• /
• 제21권5호
• /
• pp.383-392
• /
• 2005

Vegetation classes, especially grass and tree classes, are often confused in classification when conventional spectral pattern recognition techniques are used to classify urban areas. This paper reports on a study to improve the classification results by using an automated process of considering height information in separating urban vegetation classes, specifically tree and grass, using three-band, high-spatial resolution, digital aerial imagery. Height information was derived photogrammetrically from stereo pair imagery using cross correlation image matching to estimate differential parallax for vegetation pixels. A threshold value of differential parallax was used to assess whether the original class was correct. The average increase in overall accuracy for three test stereo pairs was $7.8\%$, and detailed examination showed that pixels reclassified as grass improved the overall accuracy more than pixels reclassified as tree. Visual examination and statistical accuracy assessment of four test areas showed improvement in vegetation classification with the increase in accuracy ranging from $3.7\%\;to\;18.1\%$. Vegetation classification can, in fact, be improved by adding height information to the classification procedure.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1991년도 한국자동제어학술회의논문집(국내학술편); KOEX, Seoul; 22-24 Oct. 1991
• /
• pp.1175-1180
• /
• 1991

The purpose of this paper is to improve the matching accuracy in identifying corresponding points in the area-based matching for the processing of stereo vision. For the selection of window size, a new method is proposed based on frequency domain analysis. The effectiveness of the proposed method is confirmed through a series of experiments. To overcome disproportionate distortion in stereo image pair, a new matching method using the warped window is also proposed. In the algorithm, the window is warped according to imaging geometry. Experiments on a synthetic image show that the matching accuracy is improved by 14.1% and 4.2% over the rectangular window method and image warping method each.

• 대한기계학회논문집
• /
• 제19권6호
• /
• pp.1431-1438
• /
• 1995

In this paper, we present a multiple-view stereo matching method in case of moving in the direction of optical axis with stereo camera. Also we analyze the obtainable depth precision to show that multiple-view stereo increases the virtual baseline with single-view stereo. This method decides candidate points for correspondence in each image pair and then search for the correct combinations of correspondences among them using the geometrical consistency they must satisfy. Adantages of this method are capability in increasing the accuracy in matching by using the multiple stereo images and less computation due to local processing. This method computes 3-D depth by averaging the depth obtained in each multiple-view stereo. We show that the resulting depth has more precision than depth obtainable by each independent stereo when the position of image feature is uncertain due to image noise. This paper first defines a multipleview stereo agorithm in case of moving in the direction of optical axis with stereo camera and analyze the obtainable precision of computed depth. Then we represent the effect of removing the incorrect matching candidate and precision enhancement with experimental result.

• 제어로봇시스템학회논문지
• /
• 제11권5호
• /
• pp.418-422
• /
• 2005

In this paper, a novel algorithm capable of detecting generic obstacles on a flat surface is proposed. The algorithm fundamentally exploits a distortion phenomena taken place in remapping process of original stereo images to a virtual top-view. Based on the distortion phenomena, we construct stereo polar histograms of edge maps, detect peaks on them, and search for matched peaks on both histograms using a Dynamic Programming (DP). Eventually, the searched corresponding peaks lead to estimate obstacles' positions. The advantages of the proposed algorithm are that it is not largely affected by an intensity difference between a pair of stereo images and does not depend on the typical stereo matching methodologies. Furthermore, the algorithm identifies the obstacles' positions quite robustly.

• 제어로봇시스템학회논문지
• /
• 제17권5호
• /
• pp.397-404
• /
• 2011

This paper proposes an algorithm capable of extracting corresponding points between a pair of stereo images based on dynamic programming. The purpose of extracting the corresponding points is to provide the stereo disparity data to a road-slope estimation algorithm with high accuracy and in real-time. As the road-slope estimation algorithm does not require dense disparity data, the proposed stereo matching algorithm aims at extracting corresponding points accurately and quickly. In order to realize this contradictory goal, this paper exploits dynamic programming, and minimizes matching candidates using vertical components of color edges. Furthermore, the typical occlusion problem in stereo vision is solved. The proposed algorithm is proven to be effective through experiments with various images captured on the roads.

• 대한의용생체공학회:의공학회지
• /
• 제22권6호
• /
• pp.527-534
• /
• 2001

컴퓨터 보조의 3차원 수술지원시스템(CAS. Computer Assisted Surgery)의 3차원 위치검출을 위한 광학식 추적자(Optical Tracker)에서 수술도구에 장착되는 적외선 LED의 좌우 영상을 CCD 카메라로 획득하는 데 있어서 기존의 2개의 프레임그래버를 이용하는 방법을 개선하여 하나의 그래버로 획득하는 방법을 개발하였다 좌우의 영상은 칼라 프레임그래버의 색차신호로 부가되며 이에 관련된 하드웨어 및 검출 알고리즘을 개발하였다 결과적으로 본 연구에서 개발된 방법은 비용이 절감되며 좌우영상의 추출이 빠른 것이 장점이다