[go: up one dir, main page]

KR20160083619A - Vehicle Detection Method in ROI through Plural Detection Windows - Google Patents

Vehicle Detection Method in ROI through Plural Detection Windows Download PDF

Info

Publication number
KR20160083619A
KR20160083619A KR1020140195939A KR20140195939A KR20160083619A KR 20160083619 A KR20160083619 A KR 20160083619A KR 1020140195939 A KR1020140195939 A KR 1020140195939A KR 20140195939 A KR20140195939 A KR 20140195939A KR 20160083619 A KR20160083619 A KR 20160083619A
Authority
KR
South Korea
Prior art keywords
roi
distance
rois
vehicle
region
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
KR1020140195939A
Other languages
Korean (ko)
Other versions
KR101665207B1 (en
Inventor
김진욱
Original Assignee
(주)베라시스
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by (주)베라시스 filed Critical (주)베라시스
Priority to KR1020140195939A priority Critical patent/KR101665207B1/en
Publication of KR20160083619A publication Critical patent/KR20160083619A/en
Application granted granted Critical
Publication of KR101665207B1 publication Critical patent/KR101665207B1/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Traffic Control Systems (AREA)
  • Image Processing (AREA)

Abstract

The present invention relates to a method to identify a vehicle in an interesting region via detection windows and, more specifically, to a method to identify a vehicle in an interesting region via a plurality of detection windows generating the interesting region by receiving a brightness value image from a monocular camera. The method to identify a vehicle in an interesting region comprises: a first step of inputting the nearest region among regions determined as that a vehicle is located in a current input image; a second step of inputting the furthermost region among the regions determined as that the vehicle is located in the current input image; a third step of inputting the total generation numbers of the interesting region including the furthermost distance and the nearest distance; a fourth step of calculating a rate for the width and height between the furthermost distance and the nearest distance in the interesting region; a fifth step of calculating the size of the intermediate interesting regions by using the furthermost distance, the nearest distance, and the total generation number of the interesting region; and a sixth step of applying the interesting region to an algorithm.

Description

복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법{Vehicle Detection Method in ROI through Plural Detection Windows}[0001] Vehicle Detection Method in ROI through Plural Detection Windows [

본 발명은 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법에 관한 것으로, 더욱 상세하게는, 종래 고정식으로 정해진 부분을 검출하는 경우에 비해 식별할 차량을 놓치는 차량 유실을 방지할 수 있고, 최근거리와 최원거리만 소지하면 자동으로 중간 ROI 화면을 생성하여(ROI 영상 부분은 특정크기로 Resizing 됨) 계산이 빠르고 차량식별을 빠르게 할 수 있는 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법에 관한 것이다.The present invention relates to a vehicle identification method in a region of interest via a plurality of detection windows, and more particularly, to a vehicle identification method and a vehicle identification method, If the vehicle has only the distance and the shortest distance, it automatically generates an intermediate ROI image (the ROI image is resized to a specific size), and the vehicle identification method in the ROI through a plurality of detection windows, .

종래 cctv 혹은 영상카메라와 같은 촬영장치를 통해 입력되는 입력 영상에서 객체, 예를 들면 차량을 식별하는 경우에는 관심영역(ROI)의 x(관심영역의 x좌표),y(관심영역의 y좌표),w(관심영역의 폭),h(관심영역의 높이)를 직접 설정하였다. 또는 관심영역간의 비율을 정한 상태로 ROI를 계산하였다.(X coordinate of the ROI), y (y coordinate of the ROI) in the ROI, when an object, such as a vehicle, is identified in an input image input through a photographing device such as a cctv or a video camera, , w (the width of the region of interest), and h (the height of the region of interest). Or ROI was calculated with the ratio between the regions of interest fixed.

따라서 종래에는 ROI의 영역이 제한적이어서 관심영역을 통해 객체인 차량을 식별하기가 어려운 문제점이 있었다.Accordingly, in the conventional art, there is a problem that it is difficult to identify the vehicle, which is an object, through the region of interest because the region of the ROI is limited.

[선행기술문헌][Prior Art Literature]

대한민국 공개특허공보 제10-2011-0030938호(2011.03.24공개)(발명의 명칭: 원근 면을 이용한 객체 검출 방법 및 장치)
Korean Patent Laid-Open Publication No. 10-2011-0030938 (published on Mar. 24, 2011) (Title: Method and apparatus for object detection using perspective plane)

본 발명의 목적은 상기한 바와 같은 종래의 문제점을 해결하고자 제안한 것으로,최근거리 ROI와 최원거리 ROI를 이용하여 관심영역이 자동으로 생성되는 새로운 형태의 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법을 제공하는데 있다.It is an object of the present invention to solve the above-mentioned problems of the related art, and it is an object of the present invention to provide a method and apparatus for identifying a vehicle in a region of interest through a plurality of detection windows of a new type in which a ROI is automatically generated using a distance ROI and a ROI Method.

본 발명의 또 다른 목적은 전체적인 ROI의 형태나 개수의 변경이 매우 용이한 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법을 제공하는데 있다.
It is still another object of the present invention to provide a vehicle identification method in a region of interest through a plurality of detection windows in which the form or number of ROIs can be easily changed.

상기한 바와 같은 목적을 달성하기 위한 본 발명의 바람직한 실시예에 따르면,According to a preferred embodiment of the present invention,

복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법은A vehicle identification method in a region of interest via a plurality of detection windows

단안 카메라로부터 밝기값 영상을 입력받아 관심영역을 생성하는 복수개의 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법에 있어서,A method of identifying a vehicle in a region of interest through a plurality of detection windows that receive a brightness value image from a monocular camera and generate a region of interest,

현재 입력 영상에서 차량이 위치한다고 판단된 영역들 중 가장 가까운 쪽의 영역을 입력하는 제1 단계,A first step of inputting a region of the current input image,

현재 입력 영상에서 차량이 위치한다고 판단된 영역들 중 가장 먼 쪽의 영역을 입력하는 제2 단계,A second step of inputting an area farthest from the areas where the vehicle is determined to be located in the current input image,

최원거리와 최근거리를 포함한 관심영역(ROI)의 총 생성 개수 입력하는 제3 단계,A third step of inputting a total number of generated ROIs including the best distance and the latest distance,

최원거리와 최근거리 ROI간의 폭과 높이에 대한 비율을 계산하는 제4 단계,A fourth step of calculating a ratio to the width and height between the best distance and the latest distance ROI,

최원거리와 최근거리, ROI의 총 생성개수를 이용하여 중간 ROI들의 크기를 계산하는 제5 단계,A fifth step of calculating the size of the intermediate ROIs using the best distance, the latest distance, and the total number of generated ROIs,

ROI를 알고리즘에 적용하는 제6 단계를 포함한다.And a sixth step of applying the ROI to the algorithm.

바람직하게는,Preferably,

상기 제4 단계에서 폭과 높이는 각각 수학식 1과 수학식 2에 의해 계산되는 것을 특징으로 한다.In the fourth step, the width and height are calculated by Equations (1) and (2), respectively.

(수학식 1)(1)

Figure pat00001

Figure pat00001

(수학식 2) (2)

Figure pat00002
Figure pat00002

또한 바람직하게는,Also preferably,

상기 제5 단계에서의 중간 ROI들의 크기는 x,y,폭, 높이로서, 각각 수학식 3, 수학식 4, 수학식 5, 및 수학식 6에 의해 계산되는 것을 특징으로 한다.The sizes of the intermediate ROIs in the fifth step are x, y, width, and height, and are calculated by Equations (3), (4), (5), and (6), respectively.

(수학식 3)(3)

Figure pat00003
Figure pat00003

(수학식 4)(4)

Figure pat00004
Figure pat00004

(수학식 5)(5)

Figure pat00005
Figure pat00005

(수학식 6)(6)

Figure pat00006
Figure pat00006

또한 바람직하게는 상기 제6 단계는Preferably, in the sixth step,

ROI를 알고리즘에 적용하는 단계로서,Applying an ROI to an algorithm,

원거리, 중거리, 근거리에서 식별되는 차량의 ROI가 입력영상에서의 ROI 뿐만 아니라 입력영상 일측에 표시된 복수개의 검출윈도우 중 하나의 검출윈도우에 동시에 위치되면서 표시되도록 복수개의 검출윈도우 영상 표시 과정을 제어하는 것을 특징으로 한다.
It is possible to control a plurality of detection window image display processes such that the ROI of the vehicle identified at a long distance, the medium distance, and the close range is simultaneously displayed on the detection window of one of the plurality of detection windows displayed on one side of the input image as well as the ROI on the input image .

이상 설명한 바와 같이, 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법에 의하면, 종래 고정식으로 정해진 부분을 검출하는 경우에 비해 식별할 차량을 놓치는 차량 유실을 방지할 수 있는 효과가 있고, 최근거리와 최원거리만 소지하면 자동으로 중간 ROI 화면을 생성하여(ROI 영상 부분은 특정크기로 Resizing 됨) 계산이 빠르고 차량식별을 빠르게 할 수 있는 효과가 있다. As described above, according to the vehicle identification method in the area of interest through a plurality of detection windows according to the present invention, it is possible to prevent the vehicle loss from being missed in comparison with a case where a fixed fixed part is detected (ROI image portion is resized to a specific size), and it is possible to speed up the vehicle identification by automatically generating an intermediate ROI image when having only the recent distance and the shortest distance.

또한, 최근거리 ROI를 고정시키고 최원거리 ROI를 소실점에 연동시킬 경우, 소실점의 변경이 될 때 ROI 화면의 x, y 위치를 적합하게 변경시킬 수 있어 차량 식별을 용이하게 한다.
In addition, when the distance ROI is fixed and the ROI of the closest distance is linked to the vanishing point, the x and y positions of the ROI screen can be appropriately changed when the vanishing point is changed, thereby facilitating vehicle identification.

도 1은 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법을 나타내는 흐름도이다.
도 2는 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법의 과정을 설명하는 개념도이다.
도 3 내지 도 5는 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법의 결과화면의 일예를 나타낸 도면이다.1 is a flowchart illustrating a vehicle identification method in a region of interest through a plurality of detection windows according to the present invention.
2 is a conceptual diagram illustrating a process of a vehicle identification method in a region of interest through a plurality of detection windows according to the present invention.
3 to 5 are views showing an example of a result screen of a vehicle identification method in a region of interest through a plurality of detection windows according to the present invention.

이하 본 발명에 따른 복수개의 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법을 첨부도면을 참조로 상세히 설명한다.DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A vehicle identification method in a region of interest through a plurality of detection windows according to the present invention will be described in detail with reference to the accompanying drawings.

도 1은 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법을 나타내는 흐름도이고, 도 2는 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법의 과정을 설명하는 개념도이고, 도 3 내지 도 5는 본 발명에 따른 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법의 결과화면의 일예를 나타낸 도면이다.
FIG. 1 is a flowchart illustrating a vehicle identification method in a region of interest through a plurality of detection windows according to the present invention. FIG. 2 illustrates a process of identifying a vehicle in a region of interest through a plurality of detection windows according to the present invention And FIGS. 3 to 5 are views showing an example of a result screen of a vehicle identification method in a region of interest through a plurality of detection windows according to the present invention.

도 1 내지 도 3을 참조하면, 본 발명에 따른 복수개의 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법은1 to 3, a method for identifying a vehicle in a region of interest through a plurality of detection windows according to the present invention includes:

단안 카메라로부터 밝기값 영상을 입력받아 관심영역을 생성하는 복수개의 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법에 있어서,A method of identifying a vehicle in a region of interest through a plurality of detection windows that receive a brightness value image from a monocular camera and generate a region of interest,

영상 입력과정을 통해 영상이 입력되고(S10), 이후 입력 영상에서 위치 정보를 할당하는 과정이 진행된다(S20).The image is input through the image input process (S10), and then the process of allocating the position information from the input image is performed (S20).

위치 정보를 할당하는 과정은 세부적으로,The process of assigning location information is detailed,

최근거리 ROI 정보를 입력한다(S22). 즉, 현재 입력 영상에서 차량이 위치한다고 판단된 영역들 중 가장 가까운 쪽의 영역을 입력한다.Distance ROI information is input recently (S22). That is, the nearest one of the areas determined to be located in the current input image is input.

이후 최원거리 ROI 정보를 입력한다(S23). 즉, 현재 입력 영상에서 차량이 위치한다고 판단된 영역들 중 가장 먼 쪽의 영역을 입력한다.Then, the ROI information of the best distance is inputted (S23). That is, the farthest region among the regions determined to be located in the current input image is input.

이후, 생성할 ROI 개수를 입력한다(S24). 즉, 최원거리와 최근거리를 포함한 관심영역(ROI)의 총 생성 개수 입력한다. 총 생성 개수는 제한은 2이상 9이하가 바람직하지만 이에 제한되지는 않는다.Then, the number of ROIs to be generated is input (S24). That is, the total number of ROIs including ROI and ROI is input. The total number of generations is preferably from 2 to 9, but is not limited thereto.

이후 최원거리와 최근거리 ROI간의 폭과 높이에 대한 비율을 계산한다(S26).Then, the ratio between the best distance and the latest distance ROI is calculated (S26).

이후, ROI를 생성한다(S28). 즉, 최원거리와 최근거리, ROI의 총 생성개수를 이용하여 중간 ROI들의 크기를 계산한다.Thereafter, an ROI is generated (S28). That is, the sizes of the intermediate ROIs are calculated using the best distance, the recent distance, and the total number of generated ROIs.

이후 차량식별 알고리즘을 실행한다(S30). 즉, ROI를 알고리즘에 적용한다.
Thereafter, the vehicle identification algorithm is executed (S30). That is, the ROI is applied to the algorithm.

상기 최원거리와 최근거리 ROI간의 폭과 높이를 계산하는 과정에서, 폭과 높이는 각각 수학식 1과 수학식 2에 의해 계산된다.In the process of calculating the width and height between the maximum distance and the latest distance ROI, the width and height are calculated by Equations (1) and (2), respectively.

Figure pat00007
Figure pat00007

여기서, r = 비율, w = ROI의 width, n = ROI의 전체 개수, f = 최원거리 ROI, c = 최근거리 ROI 이다. Where r = ratio, w = width of ROI, n = total number of ROIs, f = the shortest distance ROI, c = the nearest distance ROI.

Figure pat00008
Figure pat00008

여기서, r = 비율, h = ROI의 height, n = ROI의 전체 개수, f = 최원거리 ROI, c = 최근거리 ROI이다.
Where r = ratio, h = height of ROI, n = total number of ROIs, f = ROI, and c = ROI.

또한, 중간 ROI들의 크기는 x,y,폭, 높이로서, 각각 수학식 3, 수학식 4, 수학식 5, 및 수학식 6에 의해 계산된다.The sizes of the intermediate ROIs are x, y, width, and height, and are calculated by Equations 3, 4, 5, and 6, respectively.

Figure pat00009
Figure pat00009

여기서, x = ROI의 x 좌표, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율, w = ROI의 width, j=

Figure pat00010
,x와 width에 관련된 상수이다.
Here, x = x-coordinate of ROI, i = i-th latest distance ROI, i = 1, 2, ... , n, n = total number of ROIs, c = ROI distance, r = ratio, w = width of ROI, j =
Figure pat00010
, constant related to x and width.

Figure pat00011
Figure pat00011

여기서, y = ROI의 y 좌표, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율, h = ROI의 height, k=

Figure pat00012
, y와 height에 관련된 상수이다.
Here, y = y-coordinate of ROI, i = i-th latest distance ROI, i = 1, 2, ... , n = total number of ROIs, c = ROI distance, r = ratio, h = height of ROI, k =
Figure pat00012
, y and height.

Figure pat00013
Figure pat00013

여기서, w = ROI의 width, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율이다.
Here, w = width of ROI, i = ith latest distance ROI, i = 1, 2, ... , n, n = total number of ROIs, c = ROI, r = ratios.

Figure pat00014
Figure pat00014

여기서, h = ROI의 height, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율이다.
Here, h = height of ROI, i = i-th latest distance ROI, i = 1, 2, ... , n, n = total number of ROIs, c = ROI, r = ratios.

상기 ROI를 알고리즘에 적용하는 단계는 차량 식별 알고리즘 실행단계로서,원거리, 중거리, 근거리에서 식별되는 차량의 ROI가 입력영상에서의 ROI 뿐만 아니라 입력영상 일측에 표시된 복수개의 검출윈도우 중 하나의 검출윈도우에 동시에 위치되면서 표시되도록 복수개의 검출윈도우 영상 표시 과정을 제어한다.
The step of applying the ROI to the algorithm is a step of executing a vehicle identification algorithm, wherein the ROI of the vehicle, which is distinguished at a long distance, a medium distance and a short distance, is detected not only in the ROI of the input image but also in one of the plurality of detection windows A plurality of detection window image display processes are controlled so as to be displayed while being simultaneously positioned.

도 2를 참조하면, 영상화면(S-20)에서 최원거리 ROI 화면(S-11) 및 최근거리 ROI 화면(S-19)을 설정하는 과정을 볼 수있다. 자동생성된 ROI는 복수개가(S-11,S-12,S-13,S-14,S-15,S-16,S-17,S-18,S-19) 표시되었다.Referring to FIG. 2, it can be seen that the process of setting the ROI screen S-11 and the ROI screen S-19 with the shortest distance from the image screen S-20 can be seen. S-11, S-14, S-15, S-16, S-17, S-18 and S-19.

도 3을 참조하면, 원거리 상태에서 식별되는 차량 영상의 결과화면을 볼 수 있다. 즉, 영상화면(S-20)내서의 S-12에 해당하는 부분의 우측 복수개의 검출윈도우에 S-12에 해당하는 위치에 표시된 차량 식별 영상을 볼 수 있다.
Referring to FIG. 3, a result screen of the vehicle image identified in the remote state can be viewed. That is, the vehicle identification image displayed at the position corresponding to S-12 can be seen in a plurality of detection windows on the right side of the portion corresponding to S-12 in the image screen (S-20).

도 4를 참조하면, 중거리 상태에서 식별되는 차량 영상의 결과화면을 볼 수 있다. 즉, 영상화면(S-20)내의 중거리 식별영상이 해당 영역의 우측 복수개의 검출윈도우에 해당하는 위치에 표시된 차량 식별 영상을 볼 수 있다.(도 4의 중간 부분에 보임).
Referring to FIG. 4, a result screen of a vehicle image identified in a medium-distance state can be viewed. That is, a vehicle identification image in which the middle-distance identification image in the image screen S-20 is displayed at a position corresponding to a plurality of detection windows on the right side of the corresponding area can be seen (shown in the middle part of FIG. 4).

도 5를 참조하면, 근거리 상태에서 식별되는 차량 영상의 결과화면을 볼 수 있다. 즉, 영상화면(S-20)내서의 근거리 식별영상이 해당 영역의 우측 복수개의 검출윈도우에 해당하는 위치에 표시된 차량 식별 영상을 볼 수 있다.(도 5의 맨 아래측에 보임).
Referring to FIG. 5, a result screen of the vehicle image identified in the near state can be viewed. That is, the vehicle identification image in which the near-end identification image in the image screen (S-20) is displayed at a position corresponding to the plurality of detection windows on the right side of the corresponding area can be seen (shown at the bottom of FIG.

S-20: 영상화면
S-11,S-12,S-13,S-14,S-15,S-16,S-17,S-18,S-19: ROI 영상화면S-20: Video screen
S-11, S-12, S-13, S-14, S-15, S-16, S-17,

Claims (4)

단안 카메라로부터 밝기값 영상을 입력받아 관심영역을 생성하는 복수개의 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법에 있어서,
현재 입력 영상에서 차량이 위치한다고 판단된 영역들 중 가장 가까운 쪽의 영역을 입력하는 제1 단계,
현재 입력 영상에서 차량이 위치한다고 판단된 영역들 중 가장 먼 쪽의 영역을 입력하는 제2 단계,
최원거리와 최근거리를 포함한 관심영역(ROI)의 총 생성 개수 입력하는 제3 단계,
최원거리와 최근거리 ROI간의 폭과 높이에 대한 비율을 계산하는 제4 단계,
최원거리와 최근거리, ROI의 총 생성개수를 이용하여 중간 ROI들의 크기를 계산하는 제5 단계,
ROI를 알고리즘에 적용하는 제6 단계를 포함하는 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법.A method of identifying a vehicle in a region of interest through a plurality of detection windows that receive a brightness value image from a monocular camera and generate a region of interest,
A first step of inputting a region of the current input image,
A second step of inputting an area farthest from the areas where the vehicle is determined to be located in the current input image,
A third step of inputting a total number of generated ROIs including the best distance and the latest distance,
A fourth step of calculating a ratio to the width and height between the best distance and the latest distance ROI,
A fifth step of calculating the size of the intermediate ROIs using the best distance, the latest distance, and the total number of generated ROIs,
And a sixth step of applying the ROI to the algorithm. 제 1 항에 있어서,
상기 제4 단계에서 폭과 높이는 각각 수학식 1과 수학식 2에 의해 계산되는 것을 특징으로 하는 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법.
(수학식 1)
Figure pat00015

여기서, r = 비율, w = ROI의 width, n = ROI의 전체 개수, f = 최원거리 ROI, c = 최근거리 ROI 이다.
(수학식 2)
Figure pat00016

여기서, r = 비율, h = ROI의 height, n = ROI의 전체 개수, f = 최원거리 ROI, c = 최근거리 ROI이다.The method according to claim 1,
Wherein the width and the height are calculated by Equation (1) and Equation (2), respectively, in the fourth step.
(1)
Figure pat00015

Where r = ratio, w = width of ROI, n = total number of ROIs, f = the shortest distance ROI, c = the nearest distance ROI.
(2)
Figure pat00016

Where r = ratio, h = height of ROI, n = total number of ROIs, f = ROI, and c = ROI. 제 1 항에 있어서,
상기 제5 단계에서의 중간 ROI들의 크기는 x,y,폭, 높이로서, 각각 수학식 3, 수학식 4, 수학식 5, 및 수학식 6에 의해 계산되는 것을 특징으로 하는 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법.
(수학식 3)
Figure pat00017

여기서, x = ROI의 x 좌표, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율, w = ROI의 width, j=
Figure pat00018
,x와 width에 관련된 상수이다.
(수학식 4)
Figure pat00019

여기서, y = ROI의 y 좌표, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율, h = ROI의 height, k=
Figure pat00020
, y와 height에 관련된 상수이다.
(수학식 5)
Figure pat00021

여기서, w = ROI의 width, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율이다.
(수학식 6)
Figure pat00022

여기서, h = ROI의 height, i = i번째 최근거리 ROI , i = 1,2,…,n, n = ROI의 전체 개수, c = 최근거리 ROI, r = 비율이다.The method according to claim 1,
Wherein the size of the intermediate ROIs in the fifth step is x, y, width, and height, and is calculated by Equations 3, 4, 5, and 6, respectively. A method for identifying a vehicle in a region of interest through a network.
(3)
Figure pat00017

Here, x = x-coordinate of ROI, i = i-th latest distance ROI, i = 1, 2, ... , n, n = total number of ROIs, c = ROI distance, r = ratio, w = width of ROI, j =
Figure pat00018
, constant related to x and width.
(4)
Figure pat00019

Here, y = y-coordinate of ROI, i = i-th latest distance ROI, i = 1, 2, ... , n = total number of ROIs, c = ROI distance, r = ratio, h = height of ROI, k =
Figure pat00020
, y and height.
(5)
Figure pat00021

Here, w = width of ROI, i = ith latest distance ROI, i = 1, 2, ... , n, n = total number of ROIs, c = ROI, r = ratios.
(6)
Figure pat00022

Here, h = height of ROI, i = i-th latest distance ROI, i = 1, 2, ... , n, n = total number of ROIs, c = ROI, r = ratios. 제 1 항에 있어서,
상기 제6 단계는
ROI를 알고리즘에 적용하는 단계로서,
원거리, 중거리, 근거리에서 식별되는 차량의 ROI가 입력영상에서의 ROI 뿐만 아니라 입력영상 일측에 표시된 복수개의 검출윈도우 중 하나의 검출윈도우에 동시에 위치되면서 표시되도록 복수개의 검출윈도우 영상 표시 과정을 제어하는 것을 특징으로 하는 복수개의 검출윈도우를 통한 관심영역에서의 차량식별방법.The method according to claim 1,
In the sixth step
Applying an ROI to an algorithm,
It is possible to control a plurality of detection window image display processes such that the ROI of the vehicle identified at a long distance, the medium distance, and the close range is simultaneously displayed on the detection window of one of the plurality of detection windows displayed on one side of the input image as well as the ROI on the input image The method comprising the steps < RTI ID = 0.0 > of: < / RTI >
KR1020140195939A 2014-12-31 2014-12-31 Vehicle Detection Method in ROI through Plural Detection Windows Active KR101665207B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020140195939A KR101665207B1 (en) 2014-12-31 2014-12-31 Vehicle Detection Method in ROI through Plural Detection Windows

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020140195939A KR101665207B1 (en) 2014-12-31 2014-12-31 Vehicle Detection Method in ROI through Plural Detection Windows

Publications (2)

Publication Number Publication Date
KR20160083619A true KR20160083619A (en) 2016-07-12
KR101665207B1 KR101665207B1 (en) 2016-10-14

Family

ID=56505141

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020140195939A Active KR101665207B1 (en) 2014-12-31 2014-12-31 Vehicle Detection Method in ROI through Plural Detection Windows

Country Status (1)

Country Link
KR (1) KR101665207B1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018124337A1 (en) * 2016-12-28 2018-07-05 주식회사 에이다스원 Object detection method and apparatus utilizing adaptive area of interest and discovery window
CN113506264A (en) * 2021-07-07 2021-10-15 北京航空航天大学 Road vehicle number identification method and device

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10649771B2 (en) 2017-03-31 2020-05-12 Samsung Electronics Co., Ltd. Semiconductor device
KR102235803B1 (en) 2017-03-31 2021-04-06 삼성전자주식회사 Semiconductor device

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001004368A (en) * 1999-06-16 2001-01-12 Honda Motor Co Ltd Object recognition device
KR20130020151A (en) * 2011-08-19 2013-02-27 주식회사 만도 Vehicle detection device and method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001004368A (en) * 1999-06-16 2001-01-12 Honda Motor Co Ltd Object recognition device
KR20130020151A (en) * 2011-08-19 2013-02-27 주식회사 만도 Vehicle detection device and method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018124337A1 (en) * 2016-12-28 2018-07-05 주식회사 에이다스원 Object detection method and apparatus utilizing adaptive area of interest and discovery window
CN113506264A (en) * 2021-07-07 2021-10-15 北京航空航天大学 Road vehicle number identification method and device
CN113506264B (en) * 2021-07-07 2023-08-29 北京航空航天大学 Road Vehicle Number Recognition Method and Device

Also Published As

Publication number Publication date
KR101665207B1 (en) 2016-10-14

Similar Documents

Publication Publication Date Title
KR101583947B1 (en) Apparatus and method for image defogging
US9153032B2 (en) Conversion method and apparatus with depth map generation
CN103795998B (en) Image processing method and image processing equipment
CN103731651A (en) Image processing device, image processing method, and program
US20130155050A1 (en) Refinement of Depth Maps by Fusion of Multiple Estimates
US9361668B2 (en) Method and apparatus for generating disparity map
US8600112B2 (en) Crosswalk detection device, crosswalk detection method and recording medium
US20180020205A1 (en) Stereo matching system and method of operating thereof
KR101665207B1 (en) Vehicle Detection Method in ROI through Plural Detection Windows
US9779539B2 (en) Image processing apparatus and image processing method
KR101573576B1 (en) Image processing method of around view monitoring system
JP2010147560A (en) Target tracker
US20140184754A1 (en) Method of obtaining depth information and display apparatus
CN106251348B (en) An adaptive multi-cue fusion background subtraction method for depth cameras
US20140064608A1 (en) Method of transforming stereoscopic image and recording medium storing the same
US9344712B2 (en) Image processing device, image processing method, computer program product, and image display device
US10523928B2 (en) Method and apparatus for correcting distortion
CN104637043B (en) Pixel selecting method, device, parallax value is supported to determine method
KR20110015142A (en) How to measure the distance of an object using a stereo camera
US9936189B2 (en) Method for predicting stereoscopic depth and apparatus thereof
JP2012105019A (en) Image processing device, method, and program thereof
JP4892105B1 (en) Video processing device, video processing method, and video display device
KR20160024419A (en) System and Method for identifying stereo-scopic camera in Depth-Image-Based Rendering
EP2706748A1 (en) Stereoscopic image generation device, stereoscopic image generation method, stereoscopic image generation program
TWI702845B (en) Methods for deploying and scheduling cameras, monitoring systems and non-transitory computer-readable medium

Legal Events

Date Code Title Description
A201 Request for examination
PA0109 Patent application

St.27 status event code: A-0-1-A10-A12-nap-PA0109

PA0201 Request for examination

St.27 status event code: A-1-2-D10-D11-exm-PA0201

D13-X000 Search requested

St.27 status event code: A-1-2-D10-D13-srh-X000

D14-X000 Search report completed

St.27 status event code: A-1-2-D10-D14-srh-X000

R18-X000 Changes to party contact information recorded

St.27 status event code: A-3-3-R10-R18-oth-X000

PG1501 Laying open of application

St.27 status event code: A-1-1-Q10-Q12-nap-PG1501

E902 Notification of reason for refusal
PE0902 Notice of grounds for rejection

St.27 status event code: A-1-2-D10-D21-exm-PE0902

E13-X000 Pre-grant limitation requested

St.27 status event code: A-2-3-E10-E13-lim-X000

P11-X000 Amendment of application requested

St.27 status event code: A-2-2-P10-P11-nap-X000

P13-X000 Application amended

St.27 status event code: A-2-2-P10-P13-nap-X000

E701 Decision to grant or registration of patent right
PE0701 Decision of registration

St.27 status event code: A-1-2-D10-D22-exm-PE0701

GRNT Written decision to grant
PR0701 Registration of establishment

St.27 status event code: A-2-4-F10-F11-exm-PR0701

PR1002 Payment of registration fee

St.27 status event code: A-2-2-U10-U11-oth-PR1002

Fee payment year number: 1

PG1601 Publication of registration

St.27 status event code: A-4-4-Q10-Q13-nap-PG1601

R18-X000 Changes to party contact information recorded

St.27 status event code: A-5-5-R10-R18-oth-X000

R18-X000 Changes to party contact information recorded

St.27 status event code: A-5-5-R10-R18-oth-X000

FPAY Annual fee payment

Payment date: 20191007

Year of fee payment: 4

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 4

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 5

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 6

P22-X000 Classification modified

St.27 status event code: A-4-4-P10-P22-nap-X000

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 7

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 8

P22-X000 Classification modified

St.27 status event code: A-4-4-P10-P22-nap-X000

PN2301 Change of applicant

St.27 status event code: A-5-5-R10-R13-asn-PN2301

St.27 status event code: A-5-5-R10-R11-asn-PN2301

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 9

R18-X000 Changes to party contact information recorded

St.27 status event code: A-5-5-R10-R18-oth-X000

R18-X000 Changes to party contact information recorded

St.27 status event code: A-5-5-R10-R18-oth-X000

R18-X000 Changes to party contact information recorded

St.27 status event code: A-5-5-R10-R18-oth-X000

PR1001 Payment of annual fee

St.27 status event code: A-4-4-U10-U11-oth-PR1001

Fee payment year number: 10