http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
[I.T.S.부문] 화상처리 기반의 차선 휨 방향 측정 기술 개발
박종웅(Jong Woung Park),장경영(Kyung Young Jhang),이준웅(Jun Woong Lee) 한국자동차공학회 2001 한국자동차공학회 춘 추계 학술대회 논문집 Vol.2001 No.11_1
In this paper. a machine vision system to detect a lane direction and lane curvature is introduced. The system is based on image processing and MLP (multi-layered perceptron). For the neural net implementation, the lane directions are classified into three classes such as straight, left curved and right curved lane. The feature vector for input neurons is extracted by image processing. The MLP plays the role in pattern matching by mapping the input vector to one of three classes of lane directions. The estimated lane direction from MLP is used as the constraint for computing the lane curvature. As the image of father area from the driving vehicle is noisy, an optimal thresholding method is also presented. Using the real road images, the proposed system is proved to work for an advanced safety system for automobile
장경영(Kyung Young Jhang),박종웅(Jong Woung Park) 대한기계학회 2001 대한기계학회 춘추학술대회 Vol.2001 No.2
So far, various technologies have been researched for the car safety. Among these researches, the studies about the active safety vehicle (ASV) connecting the intelligent vehicle are most popular. The ASV must cope effectively with the change of the outside world. The sensors for these skills are CCD camera, infrared camera, ultrasonic waves, laser radar, and millimeter-wave radar. In this paper, we investigate the features and classes, and applying technologies for these sensors.