WO2014146415A1 - 人脸识别方法和设备 - Google Patents
人脸识别方法和设备 Download PDFInfo
- Publication number
- WO2014146415A1 WO2014146415A1 PCT/CN2013/084159 CN2013084159W WO2014146415A1 WO 2014146415 A1 WO2014146415 A1 WO 2014146415A1 CN 2013084159 W CN2013084159 W CN 2013084159W WO 2014146415 A1 WO2014146415 A1 WO 2014146415A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- face image
- block
- original face
- original
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012545 processing Methods 0.000 claims abstract description 50
- 238000005070 sampling Methods 0.000 claims abstract description 10
- 238000000605 extraction Methods 0.000 claims description 29
- 239000013598 vector Substances 0.000 claims description 23
- 238000007781 pre-processing Methods 0.000 claims description 9
- 238000010586 diagram Methods 0.000 description 7
- 230000000903 blocking effect Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 230000004927 fusion Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000002708 enhancing effect Effects 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 241000282414 Homo sapiens Species 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/24—Character recognition characterised by the processing or recognition method
- G06V30/248—Character recognition characterised by the processing or recognition method involving plural approaches, e.g. verification by template match; Resolving confusion among similar patterns, e.g. "O" versus "Q"
- G06V30/2504—Coarse or fine approaches, e.g. resolution of ambiguities or multiscale approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
- G06V40/173—Classification, e.g. identification face re-identification, e.g. recognising unknown faces across different face tracks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/467—Encoded features or binary features, e.g. local binary patterns [LBP]
Definitions
- the present application claims priority to Chinese Patent Application No. 201310088469.0, entitled “Face Recognition Method and Apparatus”, filed on March 19, 2013, the entire contents of which are incorporated herein by reference. .
- the present invention relates to the field of information technology, and in particular, to a face recognition method and device.
- Biometrics are intrinsic properties of human beings, and have strong self-stability and individual differences. Therefore, the use of biometrics for identity authentication is getting more and more attention. Among them, the use of face features for identity authentication is the most natural and direct means. Therefore, in-depth study of face recognition technology has important application value. Feature extraction is the core problem of face recognition technology, which is directly related to the accuracy of final face recognition.
- a plurality of neighborhood pixels are selected around each pixel in the face image to be recognized, and the gray value of the central pixel is used as a reference, and the gray value is smaller than the gray value of the intermediate pixel.
- the adjacent pixel is quantized to 0, and the adjacent pixel whose gradation value is greater than or equal to the intermediate pixel gradation value is quantized to 1.
- the values quantized by the neighborhood pixels are then concatenated in a certain direction to obtain a binary number, and further converted into a decimal number to the central pixel.
- the above operation is performed on all the pixels in the image in turn, and a local binary pattern (LBP) histogram of the image to be recognized is obtained.
- the feature vector of the LBP histogram of the image to be identified is compared with the feature vector of the LBP histogram of the pre-registered face image to complete face recognition.
- LBP local binary pattern
- the recognition method of the face recognition method is not high.
- the present invention provides a face recognition method and apparatus for solving the problem of low face recognition accuracy existing in the prior art.
- the present invention provides a face recognition method, including:
- the recognition result of the face image is obtained based on the obtained similarity of all the image blocks.
- the original face image is downloaded to obtain at least one squat-like image that is different in size from the original face image, including:
- the original face image is subjected to n squatting to obtain n squat-like images of different sizes, and the sizes of the n different-sized squat-like images are respectively l/4 m of the original face image size.
- n is a natural number
- m is a natural number less than n+1.
- the performing the blocking processing on each of the squat image and the original squad image comprises:
- Each of the squat sample image and the original face image is subjected to blocking processing according to 1/(S*4 n ) of the original face image size, where S is a natural number.
- the image processed by the partitioning is performed on the block to obtain the features of the image block, including:
- the method further includes:
- the present invention provides a face recognition device, including:
- An obtaining module configured to obtain an original face image to be identified;
- a sample module configured to perform a sampling on the original face image to obtain at least one squat-like image that is different in size from the original face image;
- a processing module configured to perform block processing on each of the squat sample image and the original face image to obtain at least two block-processed image blocks;
- An extraction module configured to perform feature extraction on the image block after the block processing, to obtain a feature of the image block
- a matching module configured to perform similarity matching on a feature of each of the image blocks and a feature of a corresponding image block of a pre-registered face image to obtain a similarity degree of each image block;
- an identification module configured to obtain a recognition result of the face image according to the obtained similarity of all the image blocks.
- the sampling module is specifically configured to: perform n times sampling on the original facial image to obtain n different sizes of lower jaw images,
- the size of the n different size squat samples is 1/4 m of the original face image size, where n is a natural number and m is a natural number less than n+1.
- the processing module is specifically configured to:
- Each of the squat sample image and the original face image is subjected to blocking processing according to 1/(S*4 n ) of the original face image size, where S is a natural number.
- the extraction module is specifically configured to:
- the sample module is also used to:
- the present invention provides a face recognition device, including: a memory and a processor, the memory is configured to store an execution instruction, when the face recognition device is running, the processor and the memory Inter-communication, the processor executing the execution instruction causes the face recognition device to perform the method as described in the first aspect above.
- the face recognition method and device provided by the present invention obtain a squat-like image different in size from the original face image by performing a squatting of the original face image to be recognized, and perform multi-size face image feature extraction, thereby improving The ability of the face image feature to describe the face image; by segmenting the image of the lower jaw and the original face image, the recognition result of the face image is obtained according to the similarity of all the image blocks obtained after the block processing, and the recognition result is improved.
- the accuracy of face image recognition is improved.
- FIG. 1 is a flow chart of an embodiment of a face recognition method provided by the present invention.
- FIG. 2 is a flowchart of still another embodiment of a face recognition method provided by the present invention.
- FIG. 3 is a schematic diagram of a partial binary mode LBP descriptor extraction process
- FIG. 4 is a schematic structural diagram of an embodiment of a face recognition device provided by the present invention.
- FIG. 5 is a schematic structural diagram of still another embodiment of a face recognition device according to the present invention.
- FIG. 1 is a flowchart of an embodiment of a face recognition method provided by the present invention.
- the execution subject of each step in the following method embodiments may specifically be various devices having a face recognition function, such as a mobile phone, a personal computer, a PAD, an access control device, and the like.
- the method can include:
- S101 Acquire an original face image to be recognized. 5102. Perform a squatting on the original face image to obtain at least one squat-like image that is different in size from the original face image;
- the original face image may be sampled one or more times, and each time the sample is downloaded, a lower jaw image having a different size from the original face image may be obtained.
- the size of the image of the lower jaw obtained after the sample is taken can be 1/n of the original face image, and n can take a natural number.
- the 2*2 squat sample means that the length of the squat-like image obtained from the lower squat is 1/2 of the length of the original face image, and the width is 1/2 of the width of the original face image; 4*4 The sample is that the length of the lower jaw image obtained by the lower jaw is 1 / 4 of the length of the original face image, and the width is 1/4 of the width of the original face image.
- any of the original face image and the lower sample image may be subjected to block processing as a primitive block.
- the original face image is a rectangle or a square, and the original face image is subjected to 2*2 and 4*4, respectively, to obtain the original face image size 1/4 and 1/16.
- the image of the jaw Then, the squatting image of the original face image size of 1/16 can be subjected to blocking processing as a primitive block.
- the original face image can be divided into 16 image blocks; the image of the original face image size of 1/4 can be divided into 4 image blocks; and the original face image size is 1/16.
- the sample image itself is an image block without further segmentation.
- image blocks of other sizes may be used as a base block.
- the original face image can be divided into 32 image blocks; the image of the original face image size of 1/4 can be divided into 8 image blocks; and the original face image size is 1/16.
- the sample image is divided into 2 image blocks.
- each image block obtained by the block processing may be extracted by using various existing methods.
- the LBP histogram of each image block can be obtained by using the local binary mode LBP descriptor extraction method. Further, each LBP histogram of each image block can also be extracted. The feature vector of the image block.
- the pre-registered face image is also obtained and stored according to the above steps.
- the original face image is divided into 16 image blocks, and the images obtained by the two samples are divided into 4 image blocks and 1 image block, respectively.
- the pre-registered face image also corresponds to 21 image blocks, and each image block corresponds to a set of feature vectors.
- the feature vectors of the 21 image blocks of the to-be-recognized face image obtained in step S104 may be similarly matched with the feature vectors of the corresponding image blocks of the pre-registered face image, respectively, to obtain each image block in the face image to be recognized. Similarity.
- the similarity of the obtained image blocks may be subjected to weighted fusion calculation to obtain a final similarity, and the recognition result of the face image is obtained according to the final similarity. Specifically, weighting the similarity of each image block in the to-be-identified face image obtained in step S105, that is, assigning a weight to each image block, and comparing the similarity of each image block with the product of the corresponding weights. Plus, the final similarity between the face image to be recognized and the face image registered in advance is obtained.
- the weight assignment may specifically assign a larger weight to the image block where the key part of the face (such as eyes, nose, mouth, etc.) is located, and the image block where the non-key part of the face (such as a cheek, etc.) is assigned a smaller weight,
- the sum of the weights corresponding to all image blocks is 1.
- the obtained final similarity exceeds the set threshold, it can be determined whether the face image to be recognized corresponds to the same person as the pre-registered face image, and the identity of the person is recognized.
- the face recognition method provided in this embodiment is also applicable to the recognition of other images based on biological features (fingerprint, iris, palm print, etc.).
- the face recognition method provided in this embodiment obtains a squat-like image different in size from the original face image by performing a squatting of the original face image to be recognized, and performs multi-size face image feature extraction, thereby improving the person.
- the accuracy of face image recognition. 2 is a flow chart of still another embodiment of a face recognition method provided by the present invention. As shown in FIG. 2, the execution body of each step in the following method embodiments may specifically be various devices having a face recognition function. Equipment, such as: mobile phones, personal computers, PAD, access control equipment, etc.
- the method can include:
- histogram equalization is an image enhancement method for enhancing the brightness and contrast of an image, improving image quality, increasing the layering of the key parts of the image, and improving the effect of image interpretation and recognition.
- the original face image is subjected to n times of squatting, and n different sizes of squat-like images are obtained, and the size of the n-sized squat-like images is respectively l/4 m of the original face image size, wherein , n is a natural number, and m is a natural number less than n+1;
- an image is a technique for reducing image resolution to display, store, and/or transmit an image.
- the size of the sample image is 1/4, 1/16 of the original face image size, that is, the size is (W/2)*(H/2), (W/4)*(H/4).
- the original face image size l/4 n , l / ( 2 * 4 n ), l / ( 3 * 4 n ), that is, the size of the smallest size of the image obtained in step S203 or Smaller size, segmentation of the underlying image and the original face image.
- the original face image size is W*H
- it is 1/4 2 of the original face image size that is, the size of the smallest size image (W/4). *(H/4)
- FIG. 3 is a schematic diagram of a partial binary mode LBP descriptor extraction process.
- a plurality of (eight as an example) neighboring pixel points are selected around each pixel in the image block to be centered.
- Pixel The gray value of the point (4) is used as a reference, and the adjacent pixel points whose gray value is smaller than the gray value of the intermediate pixel are quantized to 0, and the adjacent pixel points whose gray value is greater than or equal to the gray value of the intermediate pixel are quantized. 1; then the values quantized by the neighborhood pixels are connected in a certain direction (in the clockwise direction) to obtain an 8-bit binary number (11010011), and further converted into a decimal number (211) to the central pixel.
- the above operation is performed on all the pixels in the image block to obtain the LBP® of the image block.
- Each pixel in the figure corresponds to a decimal number (0 ⁇ 255), and the LBP descriptor extraction process of the image block is completed.
- the LBP histogram extracts the feature vector (X l ⁇ ) of the image block, (X 2 , Y 2 )
- the feature vectors are extracted separately for the 21 image blocks after the block processing.
- the pre-registered face image is also obtained and stored according to the above steps.
- n 2
- the feature vectors of the 21 image blocks of the to-be-recognized face image obtained in step S206 may be similarly matched with the feature vectors of the corresponding image blocks of the pre-registered face image, respectively, to obtain each image block in the face image to be recognized. Similarity.
- the similarity of the obtained image blocks may be subjected to weighted fusion calculation to obtain a final similarity, and the recognition result of the face image is obtained according to the final similarity.
- weighting and merging the similarity of each image block in the to-be-identified face image obtained in step S207 that is, assigning a weight to each image block, and comparing the similarity of each image block with the product of the corresponding weights. Plus, the final similarity between the face image to be recognized and the face image registered in advance is obtained.
- the weight assignment may specifically assign a larger weight to the image block where the key part of the face (such as eyes, nose, mouth, etc.) is located, and the image block where the non-key part of the face (such as a cheek, etc.) is assigned a smaller weight,
- the sum of the weights corresponding to all image blocks is 1.
- the image of the face to be recognized can be determined Whether the pre-registered face image corresponds to the same person, and realizes the identification of the person's identity.
- the method may further include: performing post-processing (for example, principal component analysis, linear discriminant analysis, and the like) on the extracted features of each image block to reduce the dimension of the feature of each image block, and enhancing the subsequent Discriminability of similarity matching.
- the face recognition method provided in this embodiment is also applicable to the recognition of other images based on biological features (fingerprint, iris, palm print, etc.).
- the face recognition method provided in this embodiment improves the effect of image interpretation and recognition by preprocessing the original face image to be recognized; by comparing the original face image, the image size is different from the original face image.
- the image of the lower jaw image is used to extract the feature of the face image of the multi-size image, which improves the ability of the face image feature to describe the face image; by segmenting the image of the lower jaw and the original face image after preprocessing, The recognition result of the face image is obtained according to the similarity of all the image blocks obtained after the block processing, and the accuracy of the face image recognition is improved.
- FIG. 4 is a schematic structural diagram of an embodiment of a face recognition device provided by the present invention. As shown in FIG.
- the face recognition device 40 in this embodiment is a specific body that performs the above-described face recognition method, and may specifically include: an acquisition module 41, a sample module 42, a processing module 43, an extraction module 44, and a matching module. 45 and identification module 46, wherein:
- the obtaining module 41 is configured to obtain an original face image to be identified
- the sample module 42 is configured to perform a squatting on the original face image to obtain at least one squat-like image that is different in size from the original face image; specifically, the method can: use the original face image to perform n times of sampling, N different sizes of squat-like images are obtained, and the size of the n-sized squat-like images is 1/4 m of the original face image size, where n is a natural number and m is a natural number less than n+1.
- an image is a technique for reducing image resolution to display, store, and/or transmit an image.
- the size of the squat image is 1/4, 1/16 of the original face image size, that is, the size is (W/2)*(H/2), (W/4)*(H/4).
- the processing module 43 is configured to perform block processing on each of the squat sample image and the original face image, Obtaining at least two block-processed image blocks; specifically, the method may be: performing block processing on each of the squat-like image and the original face image according to 1/(S*4 n ) of the original face image size, At least two block-processed image blocks, where S is a natural number.
- the processing module 43 can follow the original size of the original face image by l/4 n , l / ( 2 * 4 n ), 1 / ( 3 * 4 n ), that is, the minimum size of the sample module 42 is obtained.
- the size of the image is smaller or smaller, and the lower image and the original face image are subjected to blocking processing.
- the extracting module 44 is configured to perform feature extraction on the image block after the block processing to obtain the feature of the image block; specifically, the method may be: performing local binary mode LBP descriptor extraction on each block processed image block, and obtaining LBP histogram of each image block; According to the LBP histogram of each image block, the feature vector of each block-processed image block is extracted.
- the extracting module 44 performs LBP descriptor extraction on each image block obtained by the block processing, and obtains LBP® of each image block, and further obtains an LBP histogram of each image block according to the obtained LBP map, according to which The LBP histogram extracts the feature vector of each image block.
- LBP descriptor extraction on each image block obtained by the block processing, and obtains LBP® of each image block, and further obtains an LBP histogram of each image block according to the obtained LBP map, according to which The LBP histogram extracts the feature vector of each image block.
- the matching module 45 is configured to perform similarity matching on the extracted features of each image block with features of the corresponding image blocks of the pre-registered face image to obtain a similarity degree of each image block;
- the matching module 45 respectively matches the feature vectors of the 21 image blocks of the to-be-recognized face image extracted by the extraction module 44 with the feature vectors of the corresponding image blocks of the pre-registered face image, and obtains each of the to-be-recognized face images. The similarity of the image blocks.
- the identification module 46 is configured to obtain a recognition result of the face image according to the similarity of all the obtained image blocks.
- the recognition module 46 may perform a weighted fusion calculation on the similarity of all the obtained image blocks to obtain a final similarity, and obtain a recognition result of the face image according to the final similarity.
- the recognition model The block 46 performs weighted fusion on the similarity of each image block in the to-be-identified face image obtained by the matching module 45, that is, assigns a weight to each image block, and compares the similarity of each image block with the product of the corresponding weights. Plus, the final similarity between the face image to be recognized and the face image registered in advance is obtained.
- the weight assignment may specifically assign a larger weight to the image block where the focal part of the face (eg, eyes, nose, mouth, etc.) is located, and the image block where the non-key part of the face (eg, cheek, etc.) is assigned a smaller weight,
- the sum of the weights corresponding to all image blocks is 1.
- the identification module 46 determines whether the face image to be recognized corresponds to the same person as the pre-registered face image according to whether the obtained final similarity exceeds the set threshold, so as to realize the recognition of the identity of the person.
- the sample module 42 is further configured to: perform a histogram equalization process on the original face image before the original face image is downloaded.
- histogram equalization is an image enhancement method for enhancing the brightness and contrast of an image, improving image quality, increasing the layering of the key parts of the image, and improving the effect of image interpretation and recognition.
- the matching module 45 is further configured to: before the similarity of the extracted features of each image block with the features of the corresponding image block of the pre-registered face image, each image block extracted by the extraction module 44
- the features are post-processed (eg principal component analysis, linear discriminant analysis, etc.) to reduce the dimensions of the features of each image block and enhance the discriminability of subsequent similarity matching.
- the face recognition device 40 provided in this embodiment is also applicable to the recognition of other images based on biometrics (fingerprint, iris, palm print, etc.).
- the face recognition device improves the effect of image interpretation and recognition by preprocessing the original face image to be recognized; and by comparing the original face image, the image size is different from the original face image.
- the image of the lower jaw image is used to extract the feature of the face image of the multi-size image, which improves the ability of the face image feature to describe the face image; by segmenting the image of the lower jaw and the original face image after preprocessing, The recognition result of the face image is obtained according to the similarity of all the image blocks obtained after the block processing, and the accuracy of the face image recognition is improved.
- FIG. 5 is a schematic structural diagram of still another embodiment of a face recognition device according to the present invention. As shown in FIG.
- the face recognition device of this embodiment includes a memory 51 and a processor 52.
- the face recognition device may optionally include a module such as a camera to acquire an original face image.
- the memory 51 may contain a high speed RAM memory and may also include a non-volatile memory such as at least one disk memory.
- the memory 51 can optionally include at least one storage device.
- the memory 51 stores the execution The line instructions, when the face recognition device is running, the processor 52 communicates with the memory 51, and the processor 52 executes the instructions such that the face recognition device can perform the face recognition method provided by any of the embodiments of FIG. 1 or FIG.
- the processor in this embodiment may be an integrated circuit chip with signal processing capabilities.
- each step of the above method may be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software.
- the above processor may be a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component. .
- DSP digital signal processor
- ASIC application specific integrated circuit
- FPGA off-the-shelf programmable gate array
- the methods, steps, and logical block diagrams disclosed in the embodiments of the present invention may be implemented or carried out.
- the above processor may be a microprocessor or the above processor or any conventional processor or the like.
- the steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented as a hardware processor, or may be performed by a combination of hardware and software modules in the processor.
- the software modules can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like.
- the storage medium is located in the memory, and the processor reads the information in the memory and combines the hardware to complete the steps of the above method.
- the embodiment of the present invention further provides a chip for performing face recognition processing, and the chip may include the above processor.
- the face recognition device provided in this embodiment is also applicable to the recognition of other images based on biological features (fingerprint, iris, palm print, etc.).
- the face recognition device performs the pre-processing of the original face image to be recognized by the processor 52 executing the instruction stored in the memory 51, thereby improving the effect of image interpretation and recognition; Obtaining a squat-like image with a different size from the original face image, and extracting the feature image of the face image with multiple sizes, improving the ability of the face image feature to describe the face image; the image of the squat and the original after pre-processing
- the face image is subjected to block processing, and the recognition result of the face image is obtained according to the similarity of all the image blocks obtained by the block processing, thereby improving the accuracy of the face image recognition.
- the aforementioned program can be stored in a computer readable storage medium.
- the program when executed, performs the steps including the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
- Collating Specific Patterns (AREA)
- Image Processing (AREA)
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020147008279A KR20140127199A (ko) | 2013-03-19 | 2013-09-25 | 얼굴 인식 방법 및 장치 |
JP2015506094A JP2015513754A (ja) | 2013-03-19 | 2013-09-25 | 顔認識方法及びデバイス |
EP13836225.6A EP2835762B1 (en) | 2013-03-19 | 2013-09-25 | Face recognition method and device |
US14/444,611 US9405969B2 (en) | 2013-03-19 | 2014-07-28 | Face recognition method and device |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013100884690A CN103150561A (zh) | 2013-03-19 | 2013-03-19 | 人脸识别方法和设备 |
CN201310088469.0 | 2013-03-19 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/444,611 Continuation US9405969B2 (en) | 2013-03-19 | 2014-07-28 | Face recognition method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014146415A1 true WO2014146415A1 (zh) | 2014-09-25 |
Family
ID=48548624
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2013/084159 WO2014146415A1 (zh) | 2013-03-19 | 2013-09-25 | 人脸识别方法和设备 |
Country Status (6)
Country | Link |
---|---|
US (1) | US9405969B2 (zh) |
EP (1) | EP2835762B1 (zh) |
JP (1) | JP2015513754A (zh) |
KR (1) | KR20140127199A (zh) |
CN (1) | CN103150561A (zh) |
WO (1) | WO2014146415A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111079670A (zh) * | 2019-12-20 | 2020-04-28 | 北京百度网讯科技有限公司 | 人脸识别方法、装置、终端和介质 |
Families Citing this family (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103150561A (zh) * | 2013-03-19 | 2013-06-12 | 华为技术有限公司 | 人脸识别方法和设备 |
JP6235730B2 (ja) * | 2013-11-30 | 2017-11-22 | ベイジン センスタイム テクノロジー ディベロップメント カンパニー リミテッド | 顔を認識するための方法、システムおよびコンピュータ可読記憶媒体 |
CN104751108B (zh) * | 2013-12-31 | 2019-05-17 | 汉王科技股份有限公司 | 人脸图像识别装置和人脸图像识别方法 |
CN104899820B (zh) * | 2014-03-11 | 2018-11-20 | 腾讯科技(北京)有限公司 | 为图像添加标签的方法、系统和装置 |
KR102450971B1 (ko) * | 2015-05-08 | 2022-10-05 | 삼성전자주식회사 | 객체 인식 장치 및 방법 |
CN105631285A (zh) * | 2015-09-28 | 2016-06-01 | 宇龙计算机通信科技(深圳)有限公司 | 一种生物特征身份识别方法及装置 |
US9830506B2 (en) * | 2015-11-09 | 2017-11-28 | The United States Of America As Represented By The Secretary Of The Army | Method of apparatus for cross-modal face matching using polarimetric image data |
CN107135664B (zh) * | 2015-12-21 | 2020-09-11 | 厦门熵基科技有限公司 | 一种人脸识别的方法以及人脸识别装置 |
KR102707594B1 (ko) | 2016-11-11 | 2024-09-19 | 삼성전자주식회사 | 홍채 영역 추출 방법 및 장치 |
US10609383B2 (en) * | 2017-04-07 | 2020-03-31 | Hulu, LLC | Video compression using down-sampling patterns in two phases |
CN107133590B (zh) * | 2017-05-04 | 2018-09-21 | 上海天使印记信息科技有限公司 | 一种基于人脸图像的身份识别系统 |
US20190007536A1 (en) * | 2017-07-03 | 2019-01-03 | Essential Products, Inc. | Handheld writing implement form factor mobile device |
US10462345B2 (en) | 2017-08-11 | 2019-10-29 | Essential Products, Inc. | Deformable structure that compensates for displacement of a camera module of a camera accessory |
CN107563336A (zh) * | 2017-09-07 | 2018-01-09 | 廖海斌 | 用于名人匹配游戏的人脸相似度分析方法、装置和系统 |
CN108304775B (zh) | 2017-12-26 | 2021-06-11 | 北京市商汤科技开发有限公司 | 遥感图像识别方法、装置、存储介质以及电子设备 |
CN110059520B (zh) * | 2018-01-18 | 2024-04-09 | 京东科技控股股份有限公司 | 虹膜特征提取的方法、装置及虹膜识别系统 |
CN108921084A (zh) * | 2018-06-29 | 2018-11-30 | 努比亚技术有限公司 | 一种图像分类处理方法、移动终端及计算机可读存储介质 |
CN108985224A (zh) * | 2018-07-13 | 2018-12-11 | 江苏慧学堂系统工程有限公司 | 一种提升人脸识别准确率的方法和人脸识别系统 |
CN109543579A (zh) * | 2018-11-14 | 2019-03-29 | 杭州登虹科技有限公司 | 一种图像中目标对象的识别方法、装置和存储介质 |
US12003850B2 (en) * | 2019-01-31 | 2024-06-04 | Huawei Technologies Co., Ltd. | Method for selecting image based on burst shooting and electronic device |
CN109829448B (zh) * | 2019-03-07 | 2021-05-28 | 苏州市科远软件技术开发有限公司 | 人脸识别方法、装置及存储介质 |
CN110147727A (zh) * | 2019-04-15 | 2019-08-20 | 深圳壹账通智能科技有限公司 | 基于面部特征识别的地铁抽查方法及相关设备 |
CN110298249A (zh) * | 2019-05-29 | 2019-10-01 | 平安科技(深圳)有限公司 | 人脸识别方法、装置、终端及存储介质 |
CN110321858B (zh) * | 2019-07-08 | 2022-06-14 | 北京字节跳动网络技术有限公司 | 视频相似度确定方法、装置、电子设备及存储介质 |
CN110807139B (zh) * | 2019-10-23 | 2023-09-01 | 腾讯科技(深圳)有限公司 | 图片识别方法、装置、计算机可读存储介质和计算机设备 |
CN111507304B (zh) * | 2020-04-29 | 2023-06-27 | 广州市百果园信息技术有限公司 | 自适应刚性先验模型训练方法、人脸跟踪方法及相关装置 |
CN111814535B (zh) * | 2020-05-20 | 2024-05-07 | 平安科技(深圳)有限公司 | 掌纹图像的识别方法、装置、设备及计算机可读存储介质 |
CN112488062B (zh) * | 2020-12-18 | 2023-04-07 | 浙江大华技术股份有限公司 | 一种图像识别方法、装置、设备及介质 |
CN112906652A (zh) * | 2021-03-26 | 2021-06-04 | 建信金融科技有限责任公司 | 一种人脸图像的识别方法、装置、电子设备及存储介质 |
CN113449708B (zh) * | 2021-08-31 | 2022-01-07 | 深圳市爱深盈通信息技术有限公司 | 人脸识别方法、装置、设备终端和可读存储介质 |
CN115457644B (zh) * | 2022-11-10 | 2023-04-28 | 成都智元汇信息技术股份有限公司 | 一种基于扩展空间映射获得目标的识图方法及装置 |
CN117333928B (zh) * | 2023-12-01 | 2024-03-22 | 深圳市宗匠科技有限公司 | 一种人脸特征点检测方法、装置、电子设备及存储介质 |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020076088A1 (en) * | 2000-12-15 | 2002-06-20 | Kun-Cheng Tsai | Method of multi-level facial image recognition and system using the same |
CN1975759A (zh) * | 2006-12-15 | 2007-06-06 | 中山大学 | 一种基于结构主元分析的人脸识别方法 |
CN102136062A (zh) * | 2011-03-08 | 2011-07-27 | 西安交通大学 | 一种基于多分辨lbp的人脸检索方法 |
US20120213419A1 (en) * | 2011-02-22 | 2012-08-23 | Postech Academy-Industry Foundation | Pattern recognition method and apparatus using local binary pattern codes, and recording medium thereof |
CN102799870A (zh) * | 2012-07-13 | 2012-11-28 | 复旦大学 | 基于分块一致lbp和稀疏编码的单训练样本人脸识别方法 |
CN103150561A (zh) * | 2013-03-19 | 2013-06-12 | 华为技术有限公司 | 人脸识别方法和设备 |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000099722A (ja) * | 1998-09-22 | 2000-04-07 | Toshiba Corp | 人物顔認識装置及び人物顔認識方法 |
CN101079104A (zh) * | 2007-06-14 | 2007-11-28 | 上海交通大学 | 基于切信息的人脸识别方法 |
US8064653B2 (en) * | 2007-11-29 | 2011-11-22 | Viewdle, Inc. | Method and system of person identification by facial image |
KR101033098B1 (ko) | 2009-02-09 | 2011-05-06 | 성균관대학교산학협력단 | 실시간 얼굴 검출 장치 |
KR20100102949A (ko) | 2009-03-12 | 2010-09-27 | 한국전자통신연구원 | 블록 단위 기반 얼굴 인식 방법 및 그 장치 |
TWI453680B (zh) | 2010-10-08 | 2014-09-21 | Micro Star Int Co Ltd | 可抑制雜訊及環境影響之臉部辨識方法 |
CN102163283B (zh) * | 2011-05-25 | 2012-08-29 | 电子科技大学 | 一种基于局部三值模式的人脸特征提取方法 |
-
2013
- 2013-03-19 CN CN2013100884690A patent/CN103150561A/zh active Pending
- 2013-09-25 WO PCT/CN2013/084159 patent/WO2014146415A1/zh active Application Filing
- 2013-09-25 JP JP2015506094A patent/JP2015513754A/ja active Pending
- 2013-09-25 KR KR1020147008279A patent/KR20140127199A/ko not_active Application Discontinuation
- 2013-09-25 EP EP13836225.6A patent/EP2835762B1/en active Active
-
2014
- 2014-07-28 US US14/444,611 patent/US9405969B2/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020076088A1 (en) * | 2000-12-15 | 2002-06-20 | Kun-Cheng Tsai | Method of multi-level facial image recognition and system using the same |
CN1975759A (zh) * | 2006-12-15 | 2007-06-06 | 中山大学 | 一种基于结构主元分析的人脸识别方法 |
US20120213419A1 (en) * | 2011-02-22 | 2012-08-23 | Postech Academy-Industry Foundation | Pattern recognition method and apparatus using local binary pattern codes, and recording medium thereof |
CN102136062A (zh) * | 2011-03-08 | 2011-07-27 | 西安交通大学 | 一种基于多分辨lbp的人脸检索方法 |
CN102799870A (zh) * | 2012-07-13 | 2012-11-28 | 复旦大学 | 基于分块一致lbp和稀疏编码的单训练样本人脸识别方法 |
CN103150561A (zh) * | 2013-03-19 | 2013-06-12 | 华为技术有限公司 | 人脸识别方法和设备 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111079670A (zh) * | 2019-12-20 | 2020-04-28 | 北京百度网讯科技有限公司 | 人脸识别方法、装置、终端和介质 |
CN111079670B (zh) * | 2019-12-20 | 2023-11-03 | 北京百度网讯科技有限公司 | 人脸识别方法、装置、终端和介质 |
Also Published As
Publication number | Publication date |
---|---|
KR20140127199A (ko) | 2014-11-03 |
EP2835762A1 (en) | 2015-02-11 |
EP2835762A4 (en) | 2015-02-11 |
EP2835762B1 (en) | 2018-06-27 |
US9405969B2 (en) | 2016-08-02 |
JP2015513754A (ja) | 2015-05-14 |
CN103150561A (zh) | 2013-06-12 |
US20140334736A1 (en) | 2014-11-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2014146415A1 (zh) | 人脸识别方法和设备 | |
Lu et al. | Exploring finger vein based personal authentication for secure IoT | |
WO2015149534A1 (zh) | 基于Gabor二值模式的人脸识别方法及装置 | |
WO2021248733A1 (zh) | 应用双分支三维卷积模型的人脸活体检测系统、终端及存储介质 | |
CN109543548A (zh) | 一种人脸识别方法、装置及存储介质 | |
CN111814194B (zh) | 基于隐私保护的图像处理方法、装置和电子设备 | |
Lu et al. | Finger vein identification using polydirectional local line binary pattern | |
CN105512683A (zh) | 基于卷积神经网络的目标定位方法及装置 | |
JPWO2010044214A1 (ja) | 顔認識装置および顔認識方法 | |
CN107169479A (zh) | 基于指纹验证的智能移动设备敏感数据防护方法 | |
CN106650568B (zh) | 一种人脸识别方法及装置 | |
Zhu et al. | Detection of spoofing medium contours for face anti-spoofing | |
CN105681324B (zh) | 互联网金融交易系统及方法 | |
WO2020253304A1 (zh) | 人脸识别装置及图像处理方法、特征提取模型、存储介质 | |
WO2017113083A1 (zh) | 一种虹膜识别方法及装置 | |
Fischer et al. | A novel palm vein recognition approach based on enhanced local Gabor binary patterns histogram sequence | |
Purnapatra et al. | Presentation attack detection with advanced cnn models for noncontact-based fingerprint systems | |
Bruno et al. | Palmprint principal lines extraction | |
CN112686191A (zh) | 基于人脸三维信息的活体防伪方法、系统、终端及介质 | |
CN108182375B (zh) | 一种基于手机支付的指纹识别系统 | |
Mali et al. | Indian sign language recognition using SVM classifier | |
CN111126250A (zh) | 一种基于ptgan的行人重识别方法及装置 | |
Chowdhury et al. | Fuzzy rule based approach for face and facial feature extraction in biometric authentication | |
CN113743371B (zh) | 一种指纹识别方法及指纹识别装置 | |
Alharbi et al. | Spoofing Face Detection Using Novel Edge-Net Autoencoder for Security. |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 2013836225 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 2015506094 Country of ref document: JP Kind code of ref document: A Ref document number: 20147008279 Country of ref document: KR Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13836225 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |