• 로봇학회논문지
• /
• 제4권3호
• /
• pp.185-191
• /
• 2009

This paper presents a new shape-based algorithm using affine category shape model for object category recognition and model learning. Affine category shape model is a graph of interconnected nodes whose geometric interactions are modeled using pairwise potentials. In its learning phase, it can efficiently handle large pose variations of objects in training images by estimating 2-D homography transformation between the model and the training images. Since the pairwise potentials are defined on only relative geometric relationship betweenfeatures, the proposed matching algorithm is translation and in-plane rotation invariant and robust to affine transformation. We apply spectral matching algorithm to find feature correspondences, which are then used as initial correspondences for RANSAC algorithm. The 2-D homography transformation and the inlier correspondences which are consistent with this estimate can be efficiently estimated through RANSAC, and new correspondences also can be detected by using the estimated 2-D homography transformation. Experimental results on object category database show that the proposed algorithm is robust to pose variation of objects and provides good recognition performance.

• 제어로봇시스템학회논문지
• /
• 제17권10호
• /
• pp.983-988
• /
• 2011

This paper presents a robust method for ground plane detection in vision-based applications based on a monocular sequence of images with a non-stationary camera. The proposed method, which is based on the reliable estimation of the homography between two frames taken from the sequence, aims at designing a practical system to detect road surface from traffic scenes. The homography is computed using a feature matching approach, which often gives rise to inaccurate matches or undesirable matches from out of the ground plane. Hence, the proposed homography estimation minimizes the effects from erroneous feature matching by the evaluation of the difference between the predicted and the observed matrices. The method is successfully demonstrated for the detection of road surface performed on experiments to fill an information void area taken place from geometric transformation applied to captured images by an in-vehicle camera system.

• 제어로봇시스템학회논문지
• /
• 제15권11호
• /
• pp.1088-1095
• /
• 2009

In this paper, we describe the reorientation method of distorted road sign by using projection transformation for improving recognition rate of road sign. RSR (Road Sign Recognition) is one of the most important topics for implementing driver assistance in intelligent transportation systems using pattern recognition and vision technology. The RS (Road Sign) includes direction of road or place name, and intersection for obtaining the road information. We acquire input images from mounted camera on vehicle. However, the road signs are often appeared with rotation, skew, and distortion by perspective camera. In order to obtain the correct road sign overcoming these problems, projection transformation is used to transform from 4 points of image coordinate to 4 points of world coordinate. The 4 vertices points are obtained using the trajectory as the distance from the mass center to the boundary of the object. Then, the candidate areas of road sign are transformed from distorted image by using homography transformation matrix. Internal information of reoriented road signs is segmented with arrow and the corresponding indicated place name. Arrow area is the largest labeled one. Also, the number of group of place names equals to that of arrow heads. Characters of the road sign are segmented by using vertical and horizontal histograms, and each character is recognized by using SAD (Sum of Absolute Difference). From the experiments, the proposed method has shown the higher recognition results than the image without reorientation.

• 제어로봇시스템학회논문지
• /
• 제21권1호
• /
• pp.21-27
• /
• 2015

A conventional fire detection has been developed based on images captured from a fixed camera. However, It is difficult to apply current algorithms to a flying Quad-rotor to detect fire. To solve this problem, we propose that the fire detection algorithm can be modified for Quad-rotor using Ego-motion compensation. The proposed fire detection algorithm consists of color detection, motion detection, and fire determination using a randomness test. Color detection and randomness test are adapted similarly from an existing algorithm. However, Ego-motion compensation is adapted on motion detection for compensating the degree of Quad-rotor's motion using Planar Projective Transformation based on Optical Flow, RANSAC Algorithm, and Homography. By adapting Ego-motion compensation on the motion detection step, it has been proven that the proposed algorithm has been able to detect fires 83% of the time in hovering mode.

• 한국산업융합학회 논문집
• /
• 제24권6_2호
• /
• pp.707-713
• /
• 2021

Symptoms of foot-and-mouth disease include fever and drooling a lot around the hoof, blisters in the mouth, poor appetite, blisters around the hoof, and blisters around the hoof. Research is underway on smart barns that remotely manage these symptoms through cameras. Visible light cameras can measure the condition of livestock such as blisters, but cannot measure body temperature. On the other hand, infrared thermal imaging cameras can measure body temperature, but it is difficult to measure the condition of livestock. In this paper, we propose an object detection system using deep learning-based livestock detection using visible and infrared thermal imaging composite camera modules for preemptive response

• 한국통신학회논문지
• /
• 제40권12호
• /
• pp.2363-2371
• /
• 2015

본 논문에서는 FAST(Features from Accelerated Segment Test) 특징점 검출기와 SIFT(Scale Invariant Feature Transform) 특징점 서술자(descriptor)를 사용하여 시점 변화에 강인한 특징점 정합 기법을 제안한다. 기존의 FAST 기법은 영상의 에지 부분을 따라서 불필요하게 특징점을 많이 추출하게 되는데 이러한 단점을 주곡률(principal curvatures)을 적용하여 개선한다. 추출된 특징점을 SIFT 서술자를 통해 기술하고 시점이 다른 두 영상으부터 구해진 정합쌍에 RANSAC(RANdom SAmple Consensus) 기법을 통하여 호모그래피(homography)를 계산한다. 시점 변화에 강인한 특징점 정합을 위해서 기준 영상의 특징점들을 호모그래피 변환을 통해 변경된 좌표와 시점이 다른 영상의 특징점 좌표간의 유클리디언(Euclidean) 거리를 통해 정합쌍을 분류한다. 같은 물체나 장소에 대해 시점이 변화된 여러 영상에 대한 실험을 통해서 제안하는 정합 기법이 적은 계산량으로 기존의 특징점 정합 기법보다 우수한 성능을 보여주는 것을 확인하였다.

• 한국게임학회 논문지
• /
• 제13권5호
• /
• pp.43-54
• /
• 2013

최근 3차원 분야의 지속적인 발전으로 인해 보다 사실적이고 실감나는 영상을 경험할 수 있게 되었고 게임과 같은 다양한 응용분야에서 활용가능하게 되었다. 특히 가상환경에서의 객체들과 상호작용하며 이를 제어하는 증강현실 분야에 많은 발전을 가져왔다. 본 연구는 한 대의 카메라를 활용한 가상 양시점화 방법을 통해 3차원 공간의 객체를 제어하는 3차원 사용자 인터페이스를 제안한다. 이를 위해 임의의 두 카메라 위치 사이의 변환 정보를 담고 있는 호모그래피(homography) 행렬을 계산하고, 한 대의 카메라에서 얻은 2차원 손 좌표, 호모그래피 행렬 그리고 카메라의 투영행렬을 이용하여 3차원 좌표의 복원을 수행한다. 이러한 방법을 통해 보다 정확하고 빠른 3차원 정보를 얻을 수 있게 된다. 이는 두 대의 카메라를 동시에 구동할 때보다 연산량이 감소하여 실시간 처리에 효과적일 수 있으며 경제적인 부담도 줄일 수 있는 장점을 가지고 있다.

• 대한임베디드공학회논문지
• /
• 제17권1호
• /
• pp.41-49
• /
• 2022

This paper proposes a novel lane detection algorithm based on inverse perspective transformation and machine learning in lightweight embedded system. The inverse perspective transformation method is presented for obtaining a bird's-eye view of the scene from a perspective image to remove perspective effects. This method requires only the internal and external parameters of the camera without a homography matrix with 8 degrees of freedom (DoF) that maps the points in one image to the corresponding points in the other image. To improve the accuracy and speed of lane detection in complex road environments, machine learning algorithm that has passed the first classifier is used. Before using machine learning, we apply a meaningful first classifier to the lane detection to improve the detection speed. The first classifier is applied in the bird's-eye view image to determine lane regions. A lane region passed the first classifier is detected more accurately through machine learning. The system has been tested through the driving video of the vehicle in embedded system. The experimental results show that the proposed method works well in various road environments and meet the real-time requirements. As a result, its lane detection speed is about 3.85 times faster than edge-based lane detection, and its detection accuracy is better than edge-based lane detection.

• Smart Structures and Systems
• /
• 제30권5호
• /
• pp.521-535
• /
• 2022

Efficient management of deteriorating civil infrastructure is one of the most important research topics in many developed countries. In particular, the remote displacement measurement of bridges using linear variable differential transformers, global positioning systems, laser Doppler vibrometers, and computer vision technologies has been attempted extensively. This paper proposes a remote displacement measurement system using closed-circuit televisions (CCTVs) and a computer-vision-based method for in-situ bridge bearings having relatively large displacement due to temperature change in long term. The hardware of the system is composed of a reference target for displacement measurement, a CCTV to capture target images, a gateway to transmit images via a mobile network, and a central server to store and process transmitted images. The usage of CCTV capable of night vision capture and wireless data communication enable long-term 24-hour monitoring on wide range of bridge area. The computer vision algorithm to estimate displacement from the images involves image preprocessing for enhancing the circular features of the target, circular Hough transformation for detecting circles on the target in the whole field-of-view (FOV), and homography transformation for converting the movement of the target in the images into an actual expansion displacement. The simple target design and robust circle detection algorithm help to measure displacement using target images where the targets are far apart from each other. The proposed system is installed at the Tancheon Overpass located in Seoul, and field experiments are performed to evaluate the accuracy of circle detection and displacement measurements. The circle detection accuracy is evaluated using 28,542 images captured from 71 CCTVs installed at the testbed, and only 48 images (0.168%) fail to detect the circles on the target because of subpar imaging conditions. The accuracy of displacement measurement is evaluated using images captured for 17 days from three CCTVs; the average and root-mean-square errors are 0.10 and 0.131 mm, respectively, compared with a similar displacement measurement. The long-term operation of the system, as evaluated using 8-month data, shows high accuracy and stability of the proposed system.

• 정보과학회 논문지
• /
• 제42권10호
• /
• pp.1185-1194
• /
• 2015

핀홀 카메라 모델을 가정하는 기존 영상처리 기술의 평면 대 평면 간 기하 변환은 구면 파노라마 영상에서의 픽셀 좌표에는 적용될 수 없다. 본 논문에서는 구면 파노라마 영상과 평면 영상의 특징점정합 쌍이 주어졌을 때 두 영상에 포함된 평면 기하 변환 관계를 추정하는 방법을 제안한다. 정합된 특징점들로부터 평면 패턴의 위도 변화, 경도 변화, 회전 변화, 크기 변화 인자를 모두 구하여 기하 변환을 추정하는 것이 본 논문에서 제안하는 방법의 목적이다. 평면 영상을 구면 파노라마 영상에 투영하게 될 경우 두 번의 비선형 좌표계 변환이 포함되어 기하 변환식이 복잡하다. 제안하는 방법은 좌표 변환뿐만 아니라 변환에 내재된 각 인자들을 모두 알아낼 수 있는 것이 장점이다. 실험 결과 제안하는 방법은 약 1%의 오차 수준에서 기하 변환을 추정하였고 위도 및 회전 등 주요 변형 요인에 영향을 거의 받지 않았다.