Method, apparatus, and computer-readable storage medium for pattern recognition
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/00
G06K-009/62
출원번호
US-0242470
(2008-09-30)
등록번호
US-8625858
(2014-01-07)
우선권정보
JP-2007-259059 (2007-10-02)
발명자
/ 주소
Torii, Kan
Mitarai, Yusuke
Matsugu, Masakazu
Kato, Masami
Mori, Katsuhiko
Sato, Hiroshi
출원인 / 주소
Canon Kabushiki Kaisha
대리인 / 주소
Canon USA Inc IP Division
인용정보
피인용 횟수 :
1인용 특허 :
1
초록▼
A pattern recognition method, applicable to input information including a plurality of regions, includes obtaining a certainty at which each region of the input information includes a pattern, selecting one or more regions having a relatively high-level certainty among the plurality of regions, and
A pattern recognition method, applicable to input information including a plurality of regions, includes obtaining a certainty at which each region of the input information includes a pattern, selecting one or more regions having a relatively high-level certainty among the plurality of regions, and performing pattern detection processing on the selected region or regions.
대표청구항▼
1. A pattern recognition method applicable to a single image file including a plurality of regions, the method comprising: obtaining certainties in which each region of the single image file includes a respective one of a plural of human face patterns by a first face detector;determining a certainty
1. A pattern recognition method applicable to a single image file including a plurality of regions, the method comprising: obtaining certainties in which each region of the single image file includes a respective one of a plural of human face patterns by a first face detector;determining a certainty distribution of the certainties obtained by the first face detector;determining dynamically a threshold certainty based on the determined certainty distribution;selecting at least one combination of the region and the human face pattern, the selected combination corresponding to the certainties which are equal to or greater than the determined threshold certainty; anddetecting at least one human face image based on the selected combination by a second face detector. 2. The method according to claim 1, further comprising obtaining certainties by converting the single image file using n types of methods. 3. The method according to claim 1, wherein the single image file is image data, and wherein the method further comprises obtaining certainties by designating a first image pattern as belonging to the first classification and an n-th image pattern as belonging to the n-th classification. 4. The method according to claim 3, further comprising obtaining certainties by designating a first inclination image pattern as the first image pattern and an n-th inclination image pattern as the n-th image pattern. 5. A non-transitory computer-readable storage medium storing a program for causing a computer to perform the method as claimed in claim 1. 6. A pattern recognition method according to claim 1, wherein the single image file is an image file. 7. A pattern recognition method according to claim 1, wherein the first detector obtains a certainty of each classification in each region, and each classification is an affine transformation. 8. A pattern recognition method according to claim 7, wherein each classification of which the certainty is obtained by the first detector, corresponds to an orientation of one of the patterns with a first order of precision;the pattern detection processing by the second detector obtains the orientation of one of the patterns with a second order of precision; andthe second order of precision is more precise than the first order of precision. 9. A pattern recognition method according to claim 7, wherein each classification of which the certainty is obtained by the first detector, corresponds to a parameter of one of the patterns with a first order of precision;the pattern detection processing by the second detector obtains the parameter of one of the patterns with a second order of precision; andthe second order of precision is more precise than the first order of precision. 10. A pattern recognition method according to claim 1, wherein the first detector obtains the certainties using a ratio of pixels having a predetermined color in the each region of the single image file. 11. A pattern recognition method according to claim 1, wherein the second face detector performs a detection which is different from the first face detector, at least with respect to accuracy. 12. A non-transitory computer-readable storage medium that stores a program for causing a computer to execute the method according to claim 1. 13. A pattern recognition apparatus configured to identify a pattern in a single image file including a plurality of regions, the apparatus comprising: a first detection unit configured to obtain certainties in which each region of the single image file includes a respective one of a plural of human face patterns of first to n-th, where n is a natural number equal to or greater than 2, classifications by a first detector;a determination unit configured to determine a certainty distribution of the certainties obtained by the first detector;a second determination unit configured to dynamically determine a threshold certainty based on the determined certainty distribution;a selection unit configured to select at least one combination of the region and the human face pattern, the selected combination corresponding to the certainties which are equal to or greater than the determined threshold certainty; anda second detection unit configured to detecting at least one human face image based on the one or more combinations selected by the selection unit. 14. A method comprising: obtaining certainties in which each region of an image file includes at least one pattern by a first detector;determining a certainty distribution of the certainties obtained by the first detector;determining dynamically a threshold certainty based on the determined certainty distribution;selecting at least one region from the plurality of the regions, the selected region corresponding to the certainty which is equal to or greater than the determined threshold certainty; anddetecting at least one pattern based on the selected region by a second detector. 15. A non-transitory computer-readable storage medium that stores a program for causing a computer to execute the method according to claim 14. 16. A method comprising: obtaining certainties in which each region of an image file includes a respective one of patterns of first to n-th, where n is a natural number equal to or greater than 2, classifications by a first detector;determining a certainty distribution of the certainties obtained by the first detector;determining dynamically a threshold certainty based on the determined certainty distribution;selecting at least one classification from the plurality of classifications, the selected classification corresponds to the certainty which is equal to or greater than the determined threshold certainty; anddetecting at least one pattern based on the selected classification by a second detector. 17. A non-transitory computer-readable storage medium that stores a program for causing a computer to execute the method according to claim 16. 18. An apparatus comprising: an obtaining unit configured to obtain certainties in which each region of an image file includes at least one pattern by a first detector;a first determining unit configured to determine a certainty distribution of the certainties obtained by the first detector;a second determining unit configured to determine dynamically a threshold certainty based on the determined certainty distribution;a selecting unit configured to select at least one region from the plurality of the regions, the selected region corresponding to the certainty which is equal to or greater than the determined threshold certainty; anda detecting unit configured to detect at least one pattern based on the selected region by a second detector. 19. An apparatus comprising: an obtaining unit configured to obtain certainties in which each region of an image file includes a respective one of patterns of first to n-th, where n is a natural number equal to or greater than 2, classifications by a first face detector;a first determining unit configured to determine a certainty distribution of the certainties obtained by the first detector;a second determining unit configured to determine dynamically a threshold certainty based on the determined certainty distribution;a selecting unit configured to select at least one classification from the plurality of classifications, the selected classification corresponding to the certainty which is equal to or greater than the determined threshold certainty; anda detecting unit configured to detect at least one pattern based on the selected classification by a second detector.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (1)
Ray, Lawrence A.; Nicponski, Henry, Face detecting camera and method.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.