• The Journal of Korean Institute of Information Technology
• /
• v.16 no.12
• /
• pp.75-83
• /
• 2018

Recently, autonomous vehicles have been actively studied. Pedestrian detection and recognition technology is important in autonomous vehicles. Pedestrian detection using CNN(Convolutional Neural Netwrok), which is mainly used recently, generally shows good performance, but there is a performance degradation depending on the environment of the image. In this paper, we propose a pedestrian detection system applying long-term memory structure of hippocampal neural network based on CNN network with LGP-FL (Local Gradient Pattern-Feature Layer) added. First, change the input image to a size of $227{\times}227$. Then, the feature is extracted through a total of 5 layers of convolution layer. In the process, LGP-FL adds the LGP feature pattern and stores the high-frequency pattern in the long-term memory. In the detection process, it is possible to detect the pedestrian more accurately by detecting using the LGP feature pattern information robust to brightness and color change. A comparison of the existing methods and the proposed method confirmed the increase of detection rate of about 1~4%.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.10 no.9
• /
• pp.2280-2289
• /
• 2009

The death rate of pedestrians in car accidents in Korea is about 2.5 times higher than the average of OECD countries'. If a system that can detect pedestrians and send alarm to driver is built and reduces the rate, it is worth developing such a pedestrian detection system (PDS). Since the accident rate in which pedestrians are involved is higher at nighttime than in daytime, the adoption of nighttime PDS is being standardized by big auto companies. However, they are usually using expensive night visions or multiple sensors for their PDS. In this paper we propose a method for nighttime PDS using a monochrome visible spectrum camera. We could verify its superiority in both performance and real?time operation to existing algorithm through tests against video data taken in several different environments.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.5
• /
• pp.631-637
• /
• 2010

In this paper, We propose a method of motion interpretation of pedestrian for active traffic signal control. We detect pedestrian object in a movie of crosswalk area by using the code book method and acquire contour information. To do this stage fast, we use parallel processing based on CUDA (Compute Unified Device Architecture). And we remove shadow which causes shape distortion of objects. Shadow removed object is judged by using the hilbert scan distance whether to human or noise. If the objects are judged as a human, we analyze pedestrian objects' motion, face area feature, waiting time to decide that they have intetion to across a crosswalk for pdestrians. Traffic signal can be controlled after judgement.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.10
• /
• pp.162-170
• /
• 2013

In recent years, the interests and needs of the Pedestrian Protection System (PPS), which is mounted on the vehicle for the purpose of traffic safety improvement is increasing. In this paper, we propose a pedestrian candidate window extraction and unit cell histogram based HOG descriptor calculation methods. At pedestrian detection candidate windows extraction stage, the bright ratio of pedestrian and its circumference region, vertical edge projection, edge factor, and PCA reconstruction image are used. Dalal's HOG requires pixel based histogram calculation by Gaussian weights and trilinear interpolation on overlapping blocks, But our method performs Gaussian down-weight and computes histogram on a per-cell basis, and then the histogram is combined with the adjacent cell, so our method can be calculated faster than Dalal's method. Our PCA reconstruction error based pedestrian detection candidate window extraction method efficiently classifies background based on the difference between pedestrian's head and shoulder area. The proposed method improves detection speed compared to the conventional HOG just using image without any prior information from camera calibration or depth map obtained from stereo cameras.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.06a
• /
• pp.222-223
• /
• 2016

보행자 인식을 위한 컴퓨터 비전 알고리즘은 야간 상황과 같이 저조도 환경에서는 인식 성능이 떨어지고 있다. 이로 인하여 최근 저조도 환경에서 촬영된 영상으로 야간 상황에서 객체 인식 성능을 높이는 기법들이 연구되고 있다. 야간 환경은 주간 환경과는 다르게 광량이 적기 때문에 인간의 시각으로도 객체 인식에 어려움이 있고 일반적인 카메라로 촬영된 영상으로 객체 인식이 어렵다. 최근에는 NIR 카메라를 이용하여 촬영된 영상으로 야간 보행자 인식 알고리즘이 개발되고 있으나, 인식률과 객체 인식 가능 거리 및 범위가 한정적이다. 또한 기존의 야간 보행자 검출 기법들은 방대한 연산량이 필요하기 때문에 실시간 객체 인식이 불가능하다. 본 논문에서는 NIR 카메라로부터 촬영된 영상으로 preprocessing 후 ACF(Aggregated Channel Feature)를 이용하여 최근 연구되고 있는 카메라 움직임이 있는 야간 환경에서 보행자 인식 알고리즘을 PC 및 TK1 Board 환경에서 구현하고 객체 인식률을 높인다.

• Smart Media Journal
• /
• v.4 no.2
• /
• pp.26-33
• /
• 2015

In this paper, we propose an efficient automatic pedestrian removal system from a frame in a video sequence. It firstly finds pedestrians from the frame using a Histogram of Oriented Gradient(HOG) / Linear-Support Vector Machine(L-SVM) classifier, searches for proper background patches, and then the patches are used to replace the deleted pedestrians. Background patches are retrieved from the reference video sequence and a modified feather blender algorithm is applied to make boundaries of replaced blocks look naturally. The proposed system, is designed to automatically detect object and generate natural-looking patches, while most existing systems provide search operation in manual. In the experiment, the average PSNR of the replaced blocks is 19.246

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.9
• /
• pp.1035-1042
• /
• 2014

Object recognition is one of the key technologies of the monitoring system for the prevention of various intelligent crimes. The height is one of the physical information of a person, and it may be important information for identification of the person. In this paper, a method which can detect pedestrians from CCTV images and estimate the height of the detected objects, is proposed. In this method, GMM (Gaussian Mixture Model) method was used to separate the moving object from the background and the pedestrian was detected using the conditions such as the width-height ratio and the size of the candidate objects. The proposed method was applied to the CCTV video, and the height of the pedestrian at far-distance, middle- distance, near-distance was estimated for the same person, and the accuracy was evaluated. Experimental results showed that the proposed method can estimate the height of the pedestrian as the accuracy of 97% for the short-range, 98% for the medium-range, and more than 97% for the far-range. The image sizes for the same pedestrian are different as the position of him in the image, it is shown that the proposed algorithm can estimate the height of pedestrian for various position effectively.

• Journal of Korea Society of Industrial Information Systems
• /
• v.27 no.1
• /
• pp.11-18
• /
• 2022

The use of multi-spectral cameras is essential for day and night pedestrian detection. In this paper, a color camera and a thermal imaging infrared camera were used to detect pedestrians near a crosswalk for 24 hours at an intersection with a high risk of traffic accidents. For pedestrian detection, the YOLOv5 object detector was used, and the detection performance was improved by using color images and thermal images at the same time. The proposed system showed a high performance of 0.940 mAP in the day/night multi-spectral (color and thermal image) pedestrian dataset obtained from the actual crosswalk site.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.11a
• /
• pp.921-923
• /
• 2014

보행자 검출과정은 특징추출, 추출된 특징을 기반으로 한 학습과정, 그리고 학습된 데이터를 기반으로 한 분류과정으로 나눌 수 있다. 이들 중 연산시간이 가장 오래 걸리는 특징추출과정이다. 기존의 HOG 특징 추출은 하나의 학습 샘플 이미지에 대하여 많은 픽셀 연산이 필요하기 때문에 많은 시간이 소요되었다. 본 논문에서는 실시간 스트리밍 환경에서 이전 프레임의 HOG 특징 검출정보를 분석하여 다음 프레임에서 보행자가 존재 할 가능성이 높은 부분에 대해서만 특징을 추출한다. 이 방법으로 기존의 연구와 비교하여 인식성능에는 거의 영향을 주지 않고 인식 속도를 향상할 수 있다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.04a
• /
• pp.779-781
• /
• 2015

최근 여러 매체에서 화두가 되고 있는 자율 주행 자동차나 Advanced driver assistance systems (ADAS)과 같은 분야에서 보행자 검출 기술은 핵심 요소 기술 중에 하나로 손꼽히고 있다. 특히, 인간의 인지 부하(Cognitive Load)를 고려했을 때, 주행 중에 발생할 수 있는 모든 사건을 다룬다는 것은 매우 어렵기 때문에, 앞서 언급한 방법의 도움을 받아 도로 주행 중에 발생 될 수 있는 인명 사고율을 줄이고자 하는데 그 목적이 있다. 본 논문에서는 Integral Channel Feature를 사용하여 AdaBoost 알고리즘으로 보행자 검출을 위한 분류기를 구현하였다. 그 결과, INRIA에서 제공되는 Pedestrian dataset에서 Detection rate는 97%이상, False positive는 1%에 정도로 나타났다.