• 제목/요약/키워드: vision model

검색결과 1,320건 처리시간 0.028초

드론 스트리밍 영상 이미지 분석을 통한 실시간 산불 탐지 시스템 (Forest Fire Detection System using Drone Streaming Images)

  • Yoosin Kim
    • 한국항행학회논문지
    • /
    • 제27권5호
    • /
    • pp.685-689
    • /
    • 2023
  • The proposed system in the study aims to detect forest fires in real-time stream data received from the drone-camera. Recently, the number of wildfires has been increasing, and also the large scaled wildfires are frequent more and more. In order to prevent forest fire damage, many experiments using the drone camera and vision analysis are actively conducted, however there were many challenges, such as network speed, pre-processing, and model performance, to detect forest fires from real-time streaming data of the flying drone. Therefore, this study applied image data processing works to capture five good image frames for vision analysis from whole streaming data and then developed the object detection model based on YOLO_v2. As the result, the classification model performance of forest fire images reached upto 93% of accuracy, and the field test for the model verification detected the forest fire with about 70% accuracy.

확장칼만 필터링의 비젼시스템 모델을 이용한 물체 치수 측정에 관한 연구 (A Study on the Estimation of Object's Dimension based on the Vision System Model of Extended Kalman filtering)

  • 장완식;안힘찬;김경석
    • 비파괴검사학회지
    • /
    • 제25권2호
    • /
    • pp.110-116
    • /
    • 2005
  • 비젼시스템을 검사, 물체위치 결정 및 용접 작업등에 실시간 응용하고자할 때 CCD카메라에서 얻어진 많은 데이터를 처리해야 하기 때문에 전산처리 속도를 줄이는 것이 매우 중요하다. 또한, 비젼시스템 모델에 포함되어 있는 매개변수를 추정하는데 있어서 Newton-Raphson 방법 같은 반복적인 기법을 사용하면 많은 전산처리 시간을 필요하게 되어 실시간 응용을 어렵게 한다. 위의 문제점을 해결하기 위해 확장칼만필터링 같은 효율적인 방법이 필요하다. 확장칼만필터링 방법은 CCD카메라로부터 얻어지는 측정데이터의 불확실성을 고려할 뿐만 아니라, 순환적인 처리 기법을 사용하므로 전산처리 속도 향상을 가져온다. 이리하여 본 연구는 비젼시스템 모델에 포함된 카메라 내부 및 외부 매개변수를 설명하는 6개 매개변수 추정과 이를 이용한 물체 치수를 추정하는데 확장칼만필터링 방법을 적용하였다. 마지막으로 개발된 추정기법의 실시간 적용의 타당성을 실험을 통하여 보였다.

무인선의 비전기반 장애물 충돌 위험도 평가 (Vision-Based Obstacle Collision Risk Estimation of an Unmanned Surface Vehicle)

  • 우주현;김낙완
    • 제어로봇시스템학회논문지
    • /
    • 제21권12호
    • /
    • pp.1089-1099
    • /
    • 2015
  • This paper proposes vision-based collision risk estimation method for an unmanned surface vehicle. A robust image-processing algorithm is suggested to detect target obstacles from the vision sensor. Vision-based Target Motion Analysis (TMA) was performed to transform visual information to target motion information. In vision-based TMA, a camera model and optical flow are adopted. Collision risk was calculated by using a fuzzy estimator that uses target motion information and vision information as input variables. To validate the suggested collision risk estimation method, an unmanned surface vehicle experiment was performed.

Object Recognition using Smart Tag and Stereo Vision System on Pan-Tilt Mechanism

  • Kim, Jin-Young;Im, Chang-Jun;Lee, Sang-Won;Lee, Ho-Gil
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 제어로봇시스템학회 2005년도 ICCAS
    • /
    • pp.2379-2384
    • /
    • 2005
  • We propose a novel method for object recognition using the smart tag system with a stereo vision on a pan-tilt mechanism. We developed a smart tag which included IRED device. The smart tag is attached onto the object. We also developed a stereo vision system which pans and tilts for the object image to be the centered on each whole image view. A Stereo vision system on the pan-tilt mechanism can map the position of IRED to the robot coordinate system by using pan-tilt angles. And then, to map the size and pose of the object for the robot to coordinate the system, we used a simple model-based vision algorithm. To increase the possibility of tag-based object recognition, we implemented our approach by using as easy and simple techniques as possible.

  • PDF

An Optimal Combination of Illumination Intensity and Lens Aperture for Color Image Analysis

  • Chang, Y. C.
    • Agricultural and Biosystems Engineering
    • /
    • 제3권1호
    • /
    • pp.35-43
    • /
    • 2002
  • The spectral color resolution of an image is very important in color image analysis. Two factors influencing the spectral color resolution of an image are illumination intensity and lens aperture for a selected vision system. An optimal combination of illumination intensity and lens aperture for color image analysis was determined in the study. The method was based on a model of dynamic range defined as the absolute difference between digital values of selected foreground and background color in the image. The role of illumination intensity in machine vision was also described and a computer program for simulating the optimal combination of two factors was implemented for verifying the related algorithm. It was possible to estimate the non-saturating range of the illumination intensity (input voltage in the study) and the lens aperture by using a model of dynamic range. The method provided an optimal combination of the illumination intensity and the lens aperture, maximizing the color resolution between colors of interest in color analysis, and the estimated color resolution at the combination for a given vision system configuration.

  • PDF

캠시프트와 KLT특징 추적 알고리즘을 융합한 모바일 로봇의 영상기반 사람추적 및 추종 (A vision based people tracking and following for mobile robots using CAMSHIFT and KLT feature tracker)

  • 이상진;원문철
    • 한국멀티미디어학회논문지
    • /
    • 제17권7호
    • /
    • pp.787-796
    • /
    • 2014
  • Many mobile robot navigation methods utilize laser scanners, ultrasonic sensors, vision camera, and so on for detecting obstacles and path following. However, human utilizes only vision(e.g. eye) information for navigation. In this paper, we study a mobile robot control method based on only the camera vision. The Gaussian Mixture Model and a shadow removal technology are used to divide the foreground and the background from the camera image. The mobile robot uses a combined CAMSHIFT and KLT feature tracker algorithms based on the information of the foreground to follow a person. The algorithm is verified by experiments where a person is tracked and followed by a robot in a hallway.

형상 역공학을 통한 공정중 금형 가공물의 자동인식 (Automatic Recognition of In-Process mold Dies Based on Reverse Engineering Technology)

  • 김정권;윤길상;최진화;김동우;조명우;박균명
    • 한국공작기계학회:학술대회논문집
    • /
    • 한국공작기계학회 2003년도 추계학술대회
    • /
    • pp.420-425
    • /
    • 2003
  • Generally, reverse engineering means getting CAD data from unidentified shape using vision or 3D laser scanner system. In this paper, we studied unidentified model by machine vision based reverse engineering system to get information about in-processing model. Recently, vision technology is widely used in current factories, because it could inspect the in-process object easily, quickly, accurately. The following tasks were mainly investigated and implemented. We obtained more precise data by corning camera's distortion, compensating slit-beam error and revising acquired image. Much more, we made similar curves or surface with B-spline approximation for precision. Until now, there have been many case study of shape recognition. But it was uncompatible to apply to the field, because it had taken too many processing time and has frequent recognition failure. This paper propose recognition algorithm that prevent such errors and give applications to the field.

  • PDF

강건 스테레오 비전과 허프 변환을 이용한 캐드 기반 삼차원 물체인식 (CAD-Based 3-D Object Recognition Using the Robust Stereo Vision and Hough Transform)

  • 송인호;정성종
    • 한국정밀공학회:학술대회논문집
    • /
    • 한국정밀공학회 1997년도 추계학술대회 논문집
    • /
    • pp.500-503
    • /
    • 1997
  • In this paper, a method for recognizing 3-D objects using the 3-D Hough transform and the robust stereo vision is studied. A 3-D object is recognized through two steps; modeling step and matching step. In modeling step, features of the object are extracted by analyzing the IGES file. In matching step, the values of the sensed image are compared with those of the IGES file which is assumed to location and orientation in the 3-D Hough transform domain. Since we use the 3-D Hough transform domain of the input image directly, the sensitivity to the noise and the high computational complexity could be significantly allcv~ated. Also, the cost efficiency is improved using the robust stereo vision for obtaining depth map image which is needed for 3-D Hough transform. In order lo verify the proposed method, real telephone model is recognized. Thc results of the location and orientation of the model are presented.

  • PDF

Image-Based Visual Servoing Control of a SCARA Robot

  • Han, Sung-Hyun;Lee, Man-Hyung;Hashimoto, Hideki
    • Journal of Mechanical Science and Technology
    • /
    • 제14권7호
    • /
    • pp.782-788
    • /
    • 2000
  • In this paper, we present a new approach to visual feedback control using image-based visual servoing with stereo vision. In order to control the position and orientation of a robot with respect to an object, a new technique is proposed using binocular stereo vision. The stereo vision enables us to calculate an exact image Jacobian not only around a desired location but also at other locations. The suggested technique can guide a robot manipulator to the desired location without providing a priori knowledge such as the relative distance to the desired location or the model of an object even when the initial positioning error is large. This paper describes a model of stereo vision and how to generate feedback commands. The performance of the proposed visual servoing system is illustrated by experimental results and compared with conventional control methods for an assembly robot.

  • PDF

신경회로망을 이용한 지능형 로봇 제어 시스템 설계 (Design of an Intelligent Robot Control System Using Neural Network)

  • 정동연
    • 한국공작기계학회:학술대회논문집
    • /
    • 한국공작기계학회 2000년도 춘계학술대회논문집 - 한국공작기계학회
    • /
    • pp.182-187
    • /
    • 2000
  • In this paper, we have proposed a new approach to the design of robot vision system to develop the technology for the automatic test and assembling of precision mechanical and electronic parts for the factory automation. In order to perform real time implementation of the automatic assembling tasks in the complex processes, we have developed an intelligent control algorithm based-on neural networks control theory to enhance the precise motion control. Implementing of the automatic test tasks has been performed by the real-time vision algorithm based-on TMS320C31 DSPs. It distinguishes correctly the difference between the acceptable and unacceptable defective item through pattern recognition of parts by the developed vision algorithm. Finally, the performance of proposed robot vision system has been illustrated by experiment for the similar model of fifth cell among the twelve cell for similar model of fifth cell among the twelve cell for automatic test and assemblig in S company.

  • PDF