Moving object recognition apparatus

Patent metadata

Field	Value
Publication number	US-9704047-B2
Application number	US-201314422450-A
Country	US
Kind code	B2
Filing date	Jul 26, 2013
Priority date	Sep 26, 2012
Publication date	Jul 11, 2017
Grant date	Jul 11, 2017

Abstract

To provide a moving object recognition apparatus that earlier detects a moving object crossing a road, the moving object recognition apparatus includes a left imaging unit 101 , a right imaging unit 102 , a moving object detection unit that detects a moving object based on images imaged by the left imaging unit and the right imaging unit, wherein, suppose that non-overlap regions in which an imaging region of the left imaging unit 101 and an imaging region of the right imaging unit 102 do not overlap are first regions 203, 204 and an overlap region in which the imaging region of the left imaging unit and the imaging region of the right imaging unit overlap is a second region 205 , the moving object detection unit uses different methods of detecting the moving object between the first regions and the second region.

First claim

The invention claimed is:

1. A moving object recognition apparatus comprising:
a first imaging unit;
a second imaging unit; and
a processor configured to determine whether a moving object is present based on images captured by the first imaging unit and the second imaging unit,
wherein, suppose that a non-overlap region in which an imaging region of the first imaging unit and an imaging region of the second imaging unit do not overlap is a first region and an overlap region in which the imaging region of the first imaging unit and the imaging region of the second imaging unit overlap is a second region,
the processor differentiates a method of detecting the moving object between the first region and the second region,
calculates a distance to the moving object based on the images captured by the first imaging unit and the second imaging unit;
senses the moving object from an image of the first region; and
senses the moving object from an image of the second region based on the distance to the moving object and a sensing result of the moving object in the first region;
acquires a first image captured by the first imaging unit and a second image captured by the second imaging unit;
detects moving object candidates using time-series image information in the first region of the first image or the second image;
determines whether or not the detected moving object candidates enter the second region; and
outputs information of positions and velocities of only the moving object candidates that have been determined to enter the second region;
wherein the processor calculates an optical flow of the time-series image information by comparing images of two frames and identifying pixels respectively corresponding to the same object contained in the respective two images, and extracts parts showing different movements from movement of a background with movement of the moving object from obtained parts in which the images have changed and the moving object has moved;
wherein the movement of the background is cancelled from the obtained parts in which the images have changed and the moving object has moved, and the extracted parts having sizes within a range are employed as the pedestrian candidates; and
wherein the sizes within the range were obtained by previously calculating sizes in which objects having average sizes from child pedestrians to adult pedestrians are imaged on the screen, and only the pedestrian candidates moving from the first region to the second region are employed as the pedestrian candidates.

2. The moving object recognition apparatus according to claim 1 , wherein a determination is made whether or not the pedestrian candidates detected in the first region of the first image enter the second region from the first region in the next frame.

3. The moving object recognition apparatus according to claim 2 , wherein the positions and the velocities on the images of the pedestrian candidates detected in the first region of the first image are output and used as initial values for sensing moving pedestrians.

4. The moving object recognition apparatus according to claim 3 , wherein, in the first region of the second image, the pedestrian candidates are detected using images of the first region as the non-overlap region of the first and second imaging units of the second image imaged by the second imaging unit.

5. The moving object recognition apparatus according to claim 4 , wherein, for the sizes of the moving object, average values of the sizes of pedestrians from children to adults are previously learned and, if the moving object has the sizes within a range of the learned values, the moving object is determined as a pedestrian.

6. The moving object recognition apparatus according to claim 5 , wherein, for shapes of the moving object, if upper portions of the moving object resemble shapes of a head and shoulders of a human, the moving object is determined as the pedestrian.

7. The moving object recognition apparatus according to claim 5 ,
wherein initial values of the positions and velocities are provided to the moving object employed as the pedestrian candidates, the moving velocities of the moving object are obtained and employed as the velocities of moving pedestrians; and
wherein, if a determination is made that there is no moving pedestrian entering the second region from the first region and there are no initial values of the positions and the velocities of the moving pedestrians, the initial value of the velocities is set to zero and the velocities of the moving pedestrians are calculated.

8. The moving object recognition apparatus according to claim 1 , wherein, for the sizes of the moving object, average values of the sizes of pedestrians from children to adults are previously learned and, if the moving object has the sizes within a range of the learned values, the moving object is determined as a pedestrian.

9. A moving object recognition apparatus comprising:
a first imaging unit;
a second imaging unit; and
a processor configured to determine whether a moving object is present based on images captured by the first imaging unit and the second imaging unit,
wherein, suppose that a non-overlap region in which an imaging region of the first imaging unit and an imaging region of the second imaging unit do not overlap is a first region and an overlap region in which the imaging region of the first imaging unit and the imaging region of the second imaging unit overlap is a second region,
the processor differentiates a method of detecting the moving object between the first region and the second region,
calculates a distance to the moving object based on the images captured by the first imaging unit and the second imaging unit;
senses the moving object from an image of the first region; and
senses the moving object from an image of the second region based on the distance to the moving object and a sensing result of the moving object in the first region;
acquires a first image captured by the first imaging unit and a second image captured by the second imaging unit;
detects moving object candidates using time-series image information in the first region of the first image or the second image;
determines whether or not the detected moving object candidates enter the second region; and
outputs information of positions and velocities of only the moving object candidates that have been determined to enter the second region;
wherein the processor:
acquires distance information to the moving object;
detects pedestrian candidates by generating a distance image from the distance information and determining whether or not the moving object is a pedestrian from the distance image; and
determines whether or not there is a pedestrian moving from the first region to the second region in the pedestrian candidates;
when a determination that there is a pedestrian moving from the first region to the second region, sets information of the position and the velocity on the images of the moving object candidate as initial values of the pedestrian; and
calculates a moving velocity of the pedestrian;
wherein movement of a background is cancelled from obtained parts in which the images have changed and the moving object has moved, and extracted parts having sizes within a range are employed as the pedestrian candidates; and
wherein the sizes within the range were obtained by previously calculating sizes in which objects having average sizes from child pedestrians to adult pedestrians are imaged on the screen, and only the pedestrian candidates moving from the first region to the second region are employed as the pedestrian candidates.

10. The moving object recognition apparatus according to cl

Assignees

• Hitachi Automotive Systems Ltd

Inventors

• Shima Takeshi
• Matono Haruki
• Morizane Hiroto
• Monji Tatsuhiko

Classifications

• G08G1/166 Primary

  for active traffic, e.g. moving vehicles, pedestrians, bikes · CPC title

• G06V20/58

  Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads · CPC title

• G06K9/46

  Physics · mapped topic

• G06K2009/4666

  Physics · mapped topic

• G06K9/00805 Primary

  Physics · mapped topic