Moving object detection apparatus and method

• US 6,456,730 B1
• Filed: 06/17/1999
• Issued: 09/24/2002
• Est. Priority Date: 06/19/1998
• Status: Expired due to Term

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Taniguchi, Yasuhiro
• Primary Examiner(s): Johns, Andrew W.
• Assistant Examiner(s): AZARIAN, SEYED H
• Application Number:
  US09/334,948
• Time in Patent Office: 1,195 Days
• Field of Search: 382/107, 382/104, 382/103, 382/157, 348/222, 348/363, 348/169, 340/435, 340/904, 340/991, 340/903, 340/937
• US Class Current: 382/107
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• CPC Class Codes: G06V 20/588 Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
• CPC Class Codes: G06V 2201/08 Detecting or categorising vehicles

Abstract

A plurality of images are inputted in time series. A vanishing point detection section detects a vanishing point at which a plurality of lines substantially concentrate from each input image. A concentration line elimination section eliminates the plurality of lines from the each input image according to a position of the vanishing point. A moving object detection section detects a moving object from the each input image in which the plurality of lines are eliminated.

Claims

1. Moving object detection apparatus, comprising:

image input means for inputting a plurality of images in time series, wherein said image input means is loaded on a vehicle moving along a direction of a white line on a road, each input image including a rear scene to the direction;

vanishing point detection means for detecting a vanishing point at which a plurality of white line candidates concentrate most from each input image, each white line candidate being an edge line of high brightness value;

white line elimination means for eliminating the white line from each input image according to a position of the vanishing point; and

moving object detection means for detecting a moving object from each input image after the white line is eliminated from each input image.

2. The moving object detection apparatus according to claim 1,

wherein said vanishing point detection means detects edge points in a circumference area of the image, extends the edge points along a slope of the edge points inward the image, detects the extended edge line including other edge points as the white line candidate, and sequentially detects an edge line corresponding to the white line candidate from a next image in time series.

3. The moving object detection apparatus according to claim 2,

wherein said vanishing point detection means accumulates values corresponding to a position of the white line candidate of each input image onto a parameter space, and detects a peak point of all accumulated values from the parameter space as the vanishing point.

4. The moving object detection apparatus according to claim 1,

wherein said white line elimination means calculates a theoretical value of each pixel point along a direction from the pixel point to the vanishing point in the image, calculates an actual value of each pixel point along actual edge direction, and calculates a difference between the theoretical value and the actual value of each pixel point.

5. The moving object detection apparatus according to claim 4,

wherein said white line elimination means compares the difference of each pixel point with a threshold, and eliminates the white line consisting of the pixel points whose difference is below the threshold from the image.

6. The moving object detection apparatus according to claim 1,

wherein said moving object detection means divides the image without the white line into a plurality of blocks, selects a block including a clear edge element from the plurality of blocks, and pursuitably searches an area corresponding to the selected block from a next image in time series.

7. The moving object detection apparatus according to claim 6,

wherein said moving object detection means calculates a moving quantity between the block of the image and corresponding area of the next image, and stores the moving quantity between two images in time series for each selected block.

8. The moving object detection apparatus according to claim 7,

wherein said moving object detection means accumulates the moving quantity of each selected block for a set interval in time series, the set interval being long if the block is positioned near the vanishing point in the image, and the set interval being short if the block is positioned far from the vanishing point in the image.

9. The moving object detection apparatus according to claim 8,

wherein said moving object detection means counts a number of blocks in correspondence with each accumulated moving quantity as a histogram, a vertical axis of the histogram representing the number of blocks and a horizontal axis of the histogram representing the accumulated moving quantity.

10. The moving object detection apparatus according to claim 9,

wherein said moving object detection means detects the accumulated moving quantity corresponding to a peak value of the number of blocks from the histogram, and detects an area consisting of neighboring blocks corresponding to the detected accumulated moving quantity from the image as the moving object.

11. A moving object detection method comprising the steps of:

inputting a plurality of images in time series by a camera loaded on a vehicle moving along a direction of a white line on a road, each input image including a rear scene to the direction;

detecting a vanishing point at which a plurality of white line candidates concentrate most from each input image, each white line candidate being an edge line of high brightness value;

eliminating the white line from each input image according to a position of the vanishing point; and

detecting a moving object from each input image after the white line is eliminated from each input image.

12. The moving object detection method according to claim 11, further comprising the steps of:

detecting edge points in a circumference area of the image;

extending the edge points along a slope of the edge points inward the image;

detecting the extended edge line including other edge points as the white line candidate; and

sequentially detecting an edge line corresponding to the white line candidate from next image in time series.

13. The moving object detection method according to claim 12, further comprising the steps of:

accumulating values corresponding to a position of the white line candidate of each input image onto a parameter space; and

detecting a peak point of all accumulated values from the parameter space as the vanishing point.

14. The moving object detection method according to claim 11, further comprising the steps of:

calculating a theoretical value of each pixel point along a direction from the pixel point to the vanishing point in the image;

calculating an actual value of each pixel point along actual edge direction; and

calculating a difference between the theoretical value and the actual value of each pixel point.

15. The moving object detection method according to claim 14, further comprising the steps of:

comparing the difference of each pixel point with a threshold; and

eliminating the white line consisting of the pixel points whose difference is below the threshold from the image.

16. The moving object detection method according to claim 11, further comprising the steps of:

dividing the image without the white line into a plurality of blocks;

selecting a block including a clear edge element from the plurality of blocks; and

pursuitably searching an area corresponding to the selected block from a next image in time series.

17. The moving object detection method according to claim 16, further comprising the steps of:

calculating a moving quantity between the block of the image and a corresponding area of the next image; and

storing the moving quantity between two images in time series for each selected block.

18. The moving object detection method according to claim 17, further comprising the step of:

accumulating the moving quantity of each selected block for a set interval in time series, the set interval being long if the block is positioned near the vanishing point in the image, and the set interval being short if the block is positioned far from the vanishing point in the image.

19. The moving object detection method according to claim 18, further comprising the step of:

counting a number of blocks in correspondence with each accumulated moving quantity as a histogram, a vertical axis of the histogram representing the number of blocks and a horizontal axis of the histogram representing the accumulated moving quantity.

20. The moving object detection method according to claim 19, further comprising the steps of:

detecting the accumulated moving quantity corresponding to a peak value of the number of the blocks from the histogram; and

detecting an area consisting of neighboring blocks corresponding to the detected accumulated moving quantity from the image as the moving object.

21. A computer readable memory containing computer readable instructions, comprising:

instruction means for causing a computer to input a plurality of images in time series by a camera loaded on a vehicle moving along a direction of a white line on a road, each input image including a rear scene to the direction;

instruction means for causing a computer to detect a vanishing point at which a plurality of white line candidates concentrate most from each input image, each white line candidate being an edge line of high brightness value;

instruction means for causing a computer to eliminate the white line from each input image according to a position of the vanishing point; and

instruction means for causing a computer to detect a moving object from each input image after the white line is eliminated from each input image.

22. Moving object detection apparatus, comprising:

an image input unit configured to input a plurality of images in time series, wherein said image input unit is loaded on a vehicle moving along a direction of a white line on a road, each input image including a rear scene to the direction;

a vanishing point detection unit configured to detect a vanishing point at which a plurality of white line candidates concentrate most from each input image, each white line candidate being an edge line of high brightness value;

a white line elimination unit configured to eliminate the white line from each input image according to a position of the vanishing point; and

a moving object detection unit configured to detect a moving object from each input image after the white line is eliminated from each input image.