• 제어로봇시스템학회논문지
• /
• 제20권9호
• /
• pp.930-935
• /
• 2014

A Gabor filter is a linear filter used for edge detectionas frequency and orientation representations of Gabor filters are similar to those of the human visual system. In this thesis, we propose a pedestrian detection algorithm using a Gabor filter bank. In order to extract the features of the pedestrian, we use various image processing algorithms and data structure algorithms. First, color image segmentation is performed to consider the information of the RGB color space. Second, histogram equalization is performed to enhance the brightness of the input images. Third, convolution is performed between a Gabor filter bank and the enhanced images. Fourth, statistical values are calculated by using the integral image (summed area table) method. The calculated statistical values are used for the feature matrix of the pedestrian area. To evaluate the proposed algorithm, the INRIA pedestrian database and SVM (Support Vector Machine) are used, and we compare the proposed algorithm and the HOG (Histogram of Oriented Gradient) pedestrian detector, presentlyreferred to as the methodology of pedestrian detection algorithm. The experimental results show that the proposed algorithm is more accurate compared to the HOG pedestrian detector.

• 전기전자학회논문지
• /
• 제18권3호
• /
• pp.398-402
• /
• 2014

본 논문은 ROI-CENTRIST 기반 보행자 인식 알고리즘의 병렬처리 방식을 제안한다. 기존의 보행자 인식 방식만을 이용하여 임베디드 환경에서 보행자 인식을 실시간으로 처리하기에는 어려움이 존재한다. 이러한 문제는 기존의 알고리즘에 ROI를 적용한 방식을 병렬로 처리함으로써 해결할 수 있다. 본 논문에서 제안하는 ROI-CENTRIST 기반 보행자 인식의 병렬처리 방식은 기존의 CENTRIST 기반 보행자 인식 방식보다 약 10% 향상된 5.2 fps의 성능을 보인다.

• 방송공학회논문지
• /
• 제25권5호
• /
• pp.734-741
• /
• 2020

본 논문은 인간과 공존하고 커뮤니케이션하며, 인간에게 심리적 안전거리(사회적 거리) 침해에 따른 스트레스를 유발하지 않는 소셜로봇을 위한 새로운 보행자 회피 알고리즘을 제안한다. 보행자 모델을 새롭게 정의하기 위해 보행자의 걸음걸이 특성(직진성, 속도)에 따라 보행자를 클러스터링하며 보행자 클러스터별 사회적 거리를 정의한다. 정의된 사회적 거리를 포함하도록 보행자(장애물) 모델링을 하고, 새롭게 정의된 보행자 모델에 상용화된 장애물 회피, 경로계획 알고리즘을 적용해 통합된 주행 알고리즘을 완성한다. 새로운 알고리즘의 효과를 검증하기 위해, 상용화된 대표적 두가지 장애물회피 경로계획 알고리즘인 DWA 알고리즘과 TEB 알고리즘을 활용한다. 본 논문의 핵심 알고리즘인 새로운 보행자 모델을 적용한 경우와 적용하지 않은 경우로 구분하여 그 효용성을 평가한다. 그 결과, 새롭게 제안된 알고리즘이 이동시간의 손실 없이 보행자의 스트레스 지수를 현격하게 줄일 수 있음을 보인다.

• 자동차안전학회지
• /
• 제11권3호
• /
• pp.37-42
• /
• 2019

This paper presents an autonomous acceleration planning algorithm for pedestrian collision avoidance at urban. Various scenarios between pedestrians and a vehicle are designed to maneuver the planning algorithm. To simulate the scenarios, we analyze pedestrian's behavior and identify limitations of fusion sensors, lidar and vision camera. Acceleration is optimally determined by considering TTC (Time To Collision) and pedestrian's intention. Pedestrian's crossing intention is estimated for quick control decision to minimize full-braking situation, based on their velocity and position change. Feasibility of the proposed algorithm is verified by simulations using Carsim and Simulink, and comparisons with actual driving data.

• Journal of Multimedia Information System
• /
• 제6권1호
• /
• pp.1-6
• /
• 2019

In the field of industrial applications, the real-time performance of the target detection problem is very important. The most serious time consumption in the pedestrian detection process is the extraction phase of the candidate window. To accelerate the speed, in this paper, a fast extraction of pedestrian candidate window based on the BING (Binarized Normed Gradients) algorithm replaces the traditional sliding window scanning. The BING features are extracted with the positive and negative samples and input into the two-stage SVM (Support Vector Machine) classifier for training. The obtained BING template may include a pedestrian candidate window. The trained template is loaded during detection, and the extracted candidate windows are input into the classifier. The experimental results show that the proposed method can extract fewer candidate window and has a higher recall rate with more rapid speed than the traditional sliding window detection method, so the method improves the detection speed while maintaining the detection accuracy. In addition, the real-time requirement is satisfied.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제5권8호
• /
• pp.385-392
• /
• 2016

현대 사회에서 상업적 성공을 위해서는 상권 분석이 필요하며, 상권 분석의 요소 중에서 핵심적인 부분은 통행량이다. 통행량을 측정하기 위해서 사람이 직접 측정하는 방법이 많이 사용되고 있으나 높은 인건비와 측정 실수를 유발할 가능성이 높다. 본 논문에서는 웹캠을 통해 촬영한 이미지를 이용하여 보행자의 통행량을 측정할 수 있는 알고리즘을 제안한다. 제안하는 알고리즘은 사람 영역 탐지와 움직임 추적으로 구성되어 있다. 사람 영역 탐지에서는 움직임 영역을 추출하고 HoG 특징과 Adaboost 분류기를 이용하여 사람 영역을 탐지한다. 움직임 추적에는 멀티 레벨 매칭과 거짓 양성 제거를 이용하여 추적 및 통행량을 측정한다. 멀티 레벨 매칭은 HoG 영역에 대해 유사도 계수를 구하여 판별하는 과정, 칼만 필터를 이용하여 추정한 위치의 이미지 유사도를 계산 과정, 사람 영역 탐지에서 추출한 움직임 영역을 이용해 유사도를 계산하는 3단계 과정으로 구성되어 있다. 거짓 양성 제거는 사람 영역 탐지에서 잘못된 탐지 영역을 제거한다. 제안한 알고리즘의 성능을 분석하기 위하여 기존의 사람 영역 탐지 및 추적하는 방법과 비교 실험을 수행하였다. 그 결과 제안하는 방법은 사람 통행량 측정에서 83.6% 정확도를 보였으며, 기존 알고리즘에 비하여 11% 높은 성능을 달성하였다.

• 대한의용생체공학회:의공학회지
• /
• 제31권6호
• /
• pp.481-486
• /
• 2010

In this paper, we suggest an advanced algorithm, to recognize pedestrian/non-pedestrian using differential haar-like feature, which applies Adaboost algorithm to make a strong classification from weak classifications. First, we extract two feature vectors: horizontal haar-like feature and vertical haar-like feature. For the next, we calculate the proposed feature vector using differential haar-like method. And then, a strong classification needs to be obtained from weak classifications for composite recognition method using the differential area of horizontal and vertical haar-like. In the proposed method, we use one feature vector and one strong classification for the first stage of recognition. Based on our experiment, the proposed algorithm shows higher recognition rate compared to the traditional method for the pedestrian and non-pedestrian.

• 제어로봇시스템학회논문지
• /
• 제14권11호
• /
• pp.1189-1195
• /
• 2008

We introduce a MEMS-based pedestrian dead reckoning (PDR) system. A walking navigation algorithm for pedestrians is presented and map-matching algorithm for the navigation system based on dead reckoning (DR) is proposed. The PDR is equipped on the human body and provides the position information of pedestrians. And this is able to be used in ubiquitous sensor network (USN), U-hearth monitoring system, virtual reality (VR) and etc. The PDR detects a step using a novel technique and simultaneously estimates step length. Also an azimuth of the pedestrian is calculated using a fluxgate which is the one of magnetometers. Map-matching algorithm can be formulated to integrate the positioning data with the digital road network data. Map-matching algorithm not only enables the physical location to be identified from navigation system but also improves the positioning accuracy. However most of map-matching algorithms which are developed previously are for the car navigation system (CNS). Therefore they are not appropriate to implement to pedestrian navigation system based on DR system. In this paper, we propose walking navigation system and map-matching algorithm for PDR.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권5호
• /
• pp.2310-2325
• /
• 2016

In this paper, we provide a trajectory-generation algorithm that can identify pedestrians in real time. Typically, the contours for the extraction of pedestrians from the foreground of images are not clear due to factors including brightness and shade; furthermore, pedestrians move in different directions and interact with each other. These issues mean that the identification of pedestrians and the generation of trajectories are somewhat difficult. We propose a new method for trajectory generation regarding multiple pedestrians. The first stage of the method distinguishes between those pedestrian-blob situations that need to be merged and those that require splitting, followed by the use of trained decision trees to separate the pedestrians. The second stage generates the trajectories of each pedestrian by using the point-correspondence method; however, we introduce a new point-correspondence algorithm for which the A* search method has been modified. By using fuzzy membership functions, a heuristic evaluation of the correspondence between the blobs was also conducted. The proposed method was implemented and tested with the PETS 2009 dataset to show an effective multiple-pedestrian-tracking capability in a pedestrian-interaction environment.

• ETRI Journal
• /
• 제32권6호
• /
• pp.891-900
• /
• 2010

In this paper, a map-matching (MM) algorithm which combines an estimated position with digital road data is proposed. The presented algorithm using a virtual track is appropriate for a MEMS-based pedestrian dead reckoning (PDR) system, which can be used in mobile devices. Most of the previous MM algorithms are for car navigation systems and GPS-based navigation system, so existing MM algorithms are not appropriate for the pure DR-based pedestrian navigation system. The biggest problem of previous MM algorithms is that they cannot determine the correct road segment (link) due to the DR characteristics. In DR-based navigation system, the current position is propagated from the previous estimated position. This means that the MM result can be placed on a wrong link when MM algorithm fails to decide the correct link at once. It is a critical problem. Previous algorithms never overcome this problem because they did not consider pure DR characteristics. The MM algorithm using the virtual track is proposed to overcome this problem with improved accuracy. Performance of the proposed MM algorithm was verified by experiments.