Image recognition system

• US 6,658,150 B2
• Filed: 12/01/2000
• Issued: 12/02/2003
• Est. Priority Date: 12/02/1999
• Status: Expired due to Fees

• Current Assignee: Honda Giken Kogyo Kabushiki Kaisha (Honda Motor Company)
• Original Assignee: Honda Giken Kogyo Kabushiki Kaisha (Honda Motor Company)
• Inventors: Tsuji, Takayuki, Nagaoka, Nobuharu, Hattori, Hiroshi, Watanabe, Masahito
• Primary Examiner(s): Tran, Phuoc
• Application Number:
  US09/728,511
• Publication Number:
  US 20010002936A1
• Time in Patent Office: 1,096 Days
• Field of Search: 382/168, 382/170, 382/171, 382/162, 382/165, 382/154, 382/181, 382/203, 382/209, 382/218, 382/219, 348/47
• US Class Current: 382/218
• CPC Class Codes: G06T 2207/10012 Stereo images
• CPC Class Codes: G06T 7/97 Determining parameters from multiple pictures depth or shape recovery from multiple images G06T7/55; stereo camera calibration G06T7/85
• CPC Class Codes: G06V 10/7515 Shifting the patterns to accommodate for positional errors
• CPC Class Codes: H04N 13/239 using two 2D image sensors having a relative position equal to or related to the interocular distance H04N13/243 takes precedence
• CPC Class Codes: H04N 2013/0081 Depth or disparity estimation from stereoscopic image signals

Abstract

An image recognition system recognizes correspondence between object images contained respectively in two images obtained by two infrared ray cameras. One of the two images is set to a reference image and another of the two images to a comparison image. A correlation parameter indicative of a correlation between a reference pattern contained in the reference image and containing an object image and each pattern contained in the comparison image and identical in shape to the reference pattern is calculated by using a luminance value indicative of a luminance of the reference pattern and a luminance value indicative of a luminance of the each pattern contained in the comparison image. Based on the calculated correlation parameter, a pattern having a highest correlation with the reference pattern is extracted from the comparison image as a corresponding pattern which corresponds to the reference pattern. Identity between the object image contained in the reference pattern and the object image contained in the corresponding pattern is determined based on a difference between a contrast value indicative of a contrast of the reference pattern and a contrast value indicative of a contrast of the corresponding pattern.

Claims

1. An image recognition system for recognizing correspondence between object images contained respectively in two images obtained by two imaging means, the image recognition system comprising:

corresponding pattern-extracting means for setting one of the two images obtained by the two imaging means to a reference image and the other of the two images to a comparison image, calculating a correlation parameter indicative of a correlation between a reference pattern contained in the reference image and containing an object image, and each pattern contained in the comparison image and identical in shape to the reference pattern, by using a luminance value indicative of a luminance of the reference pattern and a luminance value indicative of a luminance of each pattern contained in the comparison image, and extracting, based on the calculated correlation parameter, a pattern having a highest correlation with the reference pattern from the comparison image, as a corresponding pattern which corresponds to the reference pattern; and

identity-determining means for determining identity between the object image contained in the reference pattern and an object image contained in the corresponding pattern based on a difference between a contrast value indicative of a contrast of the reference pattern and a contrast value indicative of a contrast of the corresponding pattern.

2. An image recognition system according to claim 1, wherein said identity-determining means includes contrast difference component-calculating means for calculating a correlation parameter component corresponding to the difference between the contrast value indicative of the contrast of the reference pattern and the contrast value indicative of the contrast of the corresponding pattern, as a contrast difference component, and determination means for determining identity between the object image contained in the reference pattern and the object image contained in the corresponding pattern, based on a parameter obtained by subtracting the contrast difference component from the correlation parameter concerning the corresponding pattern.

3. An image recognition system according to claim 1, wherein said identity-determining means calculates the contrast value indicative of the contrast of the reference pattern based on a difference between a luminance value indicative of a luminance of the object image contained in the reference pattern and a luminance value indicative of a luminance of a background portion within the reference pattern, and calculates the contrast value indicative of the contrast of the corresponding pattern based on a difference between a luminance value indicative of a luminance of the object image contained in the corresponding pattern and a luminance value indicative of a luminance of a background portion within the corresponding pattern.

4. An image recognition system according to claim 1, wherein said corresponding pattern-extracting means calculates overall contrast values indicative of overall contrasts of the two images, respectively, and sets one of the two images whose overall contrast value is the larger one of the calculated overall contrast values, to the reference image.

5. An image recognition system according to claim 4, wherein said corresponding pattern-extracting means calculates each of the overall contrast values indicative of the overall contrasts of the two images by the following steps:

(1) generating a histogram indicative of a frequency distribution of luminance values of each of the two images;

(2) calculating a background peak luminance value at which a portion of the histogram corresponding to a background of the each of the two images exhibits a maximum frequency;

(3) calculating an object peak luminance value at which a portion of the histogram corresponding to the object image contained in the each of the two images exhibits a maximum frequency; and

(4) calculating the overall contrast value for the each of the two images based on a difference between the object peak luminance value and the background peak luminance value of the corresponding image.

6. An image recognition system according to claim 1, wherein said identity-determining means determines that the object image contained in the reference pattern and the object image contained in the corresponding pattern are not identical when the difference between the contrast value indicative of the contrast of the reference pattern and the contrast value indicative of the contrast of the corresponding pattern is larger than a predetermined value.

7. An image recognition system according to claim 1, wherein said corresponding pattern-extracting means includes means for setting the reference pattern by clustering such that the reference pattern contains a plurality of similar objects positioned close to each other.

8. A method for recognizing correspondence between object images contained respectively in two images obtained by two imaging means, the method comprising the steps of:

setting one of the two images obtained by the two imaging means to a reference image and the other of the two images to a comparison image;

calculating a correlation parameter indicative of a correlation between a reference pattern contained in the reference image and containing an object image, and each pattern contained in the comparison image and identical in shape to the reference pattern, by using a luminance value indicative of a luminance of the reference pattern and a luminance value indicative of a luminance of each pattern contained in the comparison image;

extracting, based on the calculated correlation parameter, a pattern having a highest correlation with the reference pattern from the comparison image, as a corresponding pattern which corresponds to the reference pattern; and

determining identity between the object image contained in the reference pattern and an object image contained in the corresponding pattern based on a difference between a contrast value indicative of a contrast of the reference pattern and a contrast value indicative of a contrast of the corresponding pattern.