NZ734556B2 - Image recognition system and method - Google Patents

Image recognition system and method Download PDF

Info

Publication number
NZ734556B2
NZ734556B2 NZ734556A NZ73455615A NZ734556B2 NZ 734556 B2 NZ734556 B2 NZ 734556B2 NZ 734556 A NZ734556 A NZ 734556A NZ 73455615 A NZ73455615 A NZ 73455615A NZ 734556 B2 NZ734556 B2 NZ 734556B2
Authority
NZ
New Zealand
Prior art keywords
image
face
module
recognition
robot
Prior art date
Application number
NZ734556A
Other versions
NZ734556A (en
Inventor
Mingxiu Chen
Ningqing Liang
Hongxin Zhang
Original Assignee
Yutou Technology (Hangzhou) Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201510014262.8A external-priority patent/CN105844202A/en
Application filed by Yutou Technology (Hangzhou) Co Ltd filed Critical Yutou Technology (Hangzhou) Co Ltd
Publication of NZ734556A publication Critical patent/NZ734556A/en
Publication of NZ734556B2 publication Critical patent/NZ734556B2/en

Links

Classifications

    • G06K9/00255
    • G06K9/00268
    • G06K9/00288
    • G06K9/6269

Abstract

The invention discloses an image recognition system, comprising: a robot having an image collection module configured to drive an image collection module to collect an image in a view of the robot by the image collection drive module; a light source configured to fill the light when the image collection module collecting an image; a face detection module configured to locate a face image in an image according to the image collected by the image collection module; a face recognition module configured to implement a preprocessing for the located face image, then the preprocessed face image being compared with an image feature information of a known identity in a database, to determine an identity information and a confidence rate of the present face image. The invention can identify a face based on a fixed face pose, and it can identify a face based on a local or web server database. tion module collecting an image; a face detection module configured to locate a face image in an image according to the image collected by the image collection module; a face recognition module configured to implement a preprocessing for the located face image, then the preprocessed face image being compared with an image feature information of a known identity in a database, to determine an identity information and a confidence rate of the present face image. The invention can identify a face based on a fixed face pose, and it can identify a face based on a local or web server database.

Description

IMAGE RECOGNITION SYSTEM AND METHOD BACKGROUND OF THE INVENTION 1.Field of the Invention The present disclosure relates to a security field, more specifically, to a system and a method for image recognition for a robot system. 2.Description of the Related Art Nowadays, with the increasing requirements for the security, a great number of security systems utilize password authentication for the identification.  Notwithstanding, this kind of identification lacks in security, and people is easy to obtain the decryption method, thus the high‐level security requirement is not satisfied.  Therefore, the authentication mode through the identification of the fingerprint, iris and face is gradually adopted by the high‐level security system.  The human biological characteristics, such as face, fingerprint and iris, are innate, the uniqueness and the difficulty of being duplicated are necessary conditions for the identity authentication.  The face recognition has peculiar features compared with other types of biological recognition, these feature are as follows: non‐obligatory: the user does not need to fit the face collection device, and the face collection device can obtain the face image when the user is unconscious, thus the collection method has no “mandatory”; non‐contact: the image is collected although the user does not need to directly contact the device; concurrency: the sorting, the determination and the identification of a plurality of faces can be executed in the application scenarios; visual characteristic: people are identified by their faces, and the operation is simple, the result is intuitional, the stealthiness is good.
At present, the face recognition system generally comprises a camera configured to collect the face image, a light source configured to fill light, an assistant position system or a sign configured to prompt the face collection location, a computer (such as an embedded computer) configured to execute the face recognition software, a device configured to process or display the identification result such as a reminder light, a relay configured to open the door, and a database sheet in order to record the identification results.
The face recognition system for the security system has the following issues in the application scenarios: 1. the requirement for the pose of the face collection is fixed; 2. the light condition is sensitive and should be fixed by filling the light; 3. it does not need high speed calculating cause the calculating demand is one‐off.
SUMMARY OF THE INVENTION To resolve the shortcoming of the prior art, the invention provides an image recognition system, comprises: a robot having an image collection module, configured to drive the image collection module to collect an image in a view of the robot by the image collection drive module; a light source configured to fill the light when the image collection module collecting an image; a face detection module configured to locate a face image in the image collected by the image collection module; a face recognition module configured to implement a preprocessing for the located face image, then the preprocessed face image being compared with an image feature information of a known identity contained in a database, to determine an identity information and a confidence rate of the present face image.
Preferably, the image collection module is a high definition camera, the high definition camera is capable to obtain static image and to collect at least 30 images per second; the high definition camera is connected to the robot by a MIPI or a USB interface.
Preferably, the light source comprises an ambient light source and an infrared light source; upon utilizing the ambient light to fill the light and collecting the image by the image collection module, if the collected image can not be recognized, the light is filled by the infrared light source.
Preferably, the preprocessing comprising: executing an angle correction treatment and a light treatment (such as brightness normalization and polarized light correction) for the face image presenting in the image.
Preferably, the database comprises a local data memory module and a web server data memory module.
Preferably, the robot further comprises a sounding device connected to the database, the sounding device sends out various types of prompt tones according to a comparative result from the face recognition module.
Preferably, the image recognition system further comprises a record feedback device to record and/or feed back a comparative result from the face recognition module.
Preferably, the face recognition module implements a comparison by SVM algorithm.
A image recognition method for the above‐mentioned system, comprises the steps of: (a) collecting the image in the view of the robot by the image collection module of the robot while filling the light by a light source as the image collection module collecting the image; (b) using the face detection module to implement a location process to the face image presenting in the image  which is collected by the image collection module; (c) using the face detection module to implement a preprocessing for the face after being located, then comparing with an image feature information of a known identity contained in a database, to determine an identity information and a confidence rate of the present face image.
Preferably, in the above method, if the identity information of present face image is incompatible with an image feature information of a known identity in a database, execute the step (a) to step (c) continually.
BRIEF DESCRIPTIONS OF THE DRAWINGS Referring to the accompanying drawings, the description made by the unlimited embodiments, the disclosure and the feature, outline and advantage thereof will be more obvious.  The same signs in all drawings indicate the same portions and they are not drawn in proportion intentionally.  It illustrates the meaning of the invention.
Figure 1 shows the identity recognition system structure and the operation case thereof according to the present invention.
DETAILED DESCRIPTIONS The following description provides the details to permit a better comprehension of the invention.  However, it is obvious for the people skilled in the art that the invention can be implemented without any one or more details.  In other examples, to avoid confusion, the known technical features in the art are not described.
To understand the invention thoroughly, the following descriptions will provide detail steps and structures to explain the technical solution for the invention.  The preferred embodiment is described as follows.  However, the invention has further embodiments beyond the detailed description.
Since the invention consists in a part of the robot visual system to implement the face identification, the issues in the application of the robot should be addressed.  The issues comprise: 1 various poses of the face presenting in a view of the robot; 2. various light condition, including polarized light or no filling light source; 3. real‐time recognizing the face in the view of the robot that demanding a quick response, and the real‐time feedback implemented through a continuous recognition in accordance with the variation of the identified face.
To resolve the above issues, the embodiment provides an image recognition system, comprising: a robot having an image collection module, configured to drive the image collection module to collect an image in a view of the robot by the image collection drive module; a light source configured to fill the light when the image collection module collecting an image; a face detection module configured to locate a face image in the image collected by the image collection module; a face recognition module configured to implement a preprocessing for the located face image, then the preprocessed face image being compared with an image feature information of a known identity contained in a database, to determine an identity information and a confidence rate of the present face image.
In an embodiment of the invention, it is optional but unlimited that, the image collection module of the robot is a high definition camera.  Preferably, the high definition camera is capable to obtain static image and to collect at least 30 images per second, tofurther meet the requirement of high‐speed image collection.  For example, even if the object in the view of the robot moves fast, the invention is also able to collect the image clearly.  It is optional but unlimited that, the high definition camera is connected to the robot by a MIPI or an USB interface.  In some optional embodiments, the robot can implement a real‐time adjustment for the framing scope or angle of the image collection module througha motor.  For example, if the image collection module detects people crossing within the visual scope of it, the image collection module can implement real‐time tracking snapshot by the motor, such as moving with the object simultaneously and implementing the enlarged amplifying snapshot immediately, to improve the definition of the collected image.
In an embodiment of the invention, it is optional but unlimited that, the above light source includes an ambient light source and an infrared light source.  The ambient light source is a build‐in light source of the robot.  The advantage of the ambient light source is homogeneity of the lighting.  However the disadvantage of the ambient light source is that the luminance is low and it can not be directed.  Furthermore, the luminance can be controlled by other high‐level application, the light even be closed sometimes.  Hence, the ambient light can not meet the requirement of filling light.  Because a set of infrared light‐emitting device is added to the invention and is used to fill the light of the image, the lumination power is controlled by the image recognition system limitedly, to achieve stable filling light in various scenarios.  For example, when the ambient light source is used to fill light and to collect image by the image collection module, if the collected image can not meet the requirement of the recognition, then the invention uses the infrared light source to fill the light, thus to obtain a clear image.
In an embodiment of the invention, it is optional but unlimited that, the face detection module implements the location, that is, the face position is located in the full‐field image collected by the robot.  In the present security system, this step is unnecessary as the position is fixed.  Then preprocess the located face image by the face recognition module, the preprocessed face image is compared with an image feature information of a known identity in a database, to determine an identity information and a confidence rate of the present face image.
The face recognition module can execute the angle correction treatment and the light treatment (such as brightness normalization and polarized light correction) for the face image presented in the image.  Due to the great change of the ambient and angle of the face collection, the invention increases the recognition rate through the execution of  the angle correction treatment for the face image collected and located by the face detection module.  Meanwhile, the embedded recognition technology also can process the light of the image, to facilitate the comparison and to increase the accuracy.
In an embodiment of the invention, it is optional but unlimited that, the above database includes a local data memory module and a web server data memory module.  Based on the embedded system of the robot, the local data memory module adopts the face recognition technology according to the feature matching.  Firstly, the local data memory module makes the feature database of the face image of a known identity, secondly it extracts a same type of feature according to the face image of the real‐time collection, comparing the feature distance of the present face to the feature distance of the database face through a math distance function.  Finally, it determines the possible identity, and gives out the confidence rate.  Since the web server data memory module has more computing resources and more flexible application architectures, it adopts the face recognition technology based on a deep learning model which is configured to generate the face feature in the database, to construct the face category in the database by SVM (Support Vector Machine) algorithm or other standard classifiers.  Then it calculates the model feature according to the face image of the real‐time collection, and determines the identity and confidence rate of the face image by the classifier.  The recognition technology of the embedded feature matching supports the identification of 20‐50 people.  In a certain variation range of light and angle, the recognition accuracy rate of 20 people is greater than 90%, the recognition accuracy rate of 50 people is greater than 80%, The recognition technology on the server based on the deep learning supports the recognition from 50 people to hundreds people at least, and the recognition accuracy rate is greater than 97%.
In an embodiment of the invention, it is optional but unlimited that, the robot has a sounding device connected to the above database, the sounding device sends out various types of prompt tones according to a comparative result from the face recognition module.  For example, if the face recognition module has a correct comparison, the sounding device gets the prompt tones corresponding to the present face image in the database, such as “Hello, Mr. Chen”.  If the the regarded identification has failed after crosschecks, the greeting application still can sends out a general greeting without the identify information, such as sending out a simple “Hello” by the sounding device.  In an optional embodiment, the invention can be connected to the access control system.  If the identification succeeds, the access is allowed, otherwise if the identification fails, the access is not allowed.
In an embodiment of the invention, it is optional but unlimited that, the image recognition system provided by the invention further comprises a record feedback device to record and/or feed back a comparative result from the face recognition module.  It is an optional member, the record function and the feedback function don’t need to be implemented simultaneously.  In some scenarios, only one of them is needed.
Meanwhile, the invention also provides a recognition method by the above‐mentioned image recognition system, comprising the steps (a) collecting the image in the view of the robot by the image collection module of the robot, and filling the light through a light source as the image collection module collecting the image; (b) using the face detection module to implement a location process to the face image presenting in the image which is collected by the image collection module; (c) using the face detection module to implement a preprocessing for the face image after being located, then comparing with an image feature information of a known identity contained in a database, to determine an identity information and a confidence rate of the present face image.
If the identity information of present face image is incompatible with an image feature information of a known identity in a database, execute the step (a) to step (c) continually.
Figure 1 illustrates completely the operation process of the identity recognition system: firstly, the greeting application of the robot sends a recognition request to the image collection drive module, the image collection drive module accepts the request, and transmits the image sent from the camera to the face detection module.  The detection software truncates the located face image and executes the preprocess, and then sends the processed image to the face recognition module.  The face recognition module transmits the recognition result to the result verification module, the system executes the corresponding operation according to the determining result.  If the result is correct, the result is sent to the greeting application, and the application utilizes the sounding device to greet the user in the image of the camera according to the recognized identity; if the result is wrong, the system retransmits the recognition request to the image collection drive, and reenters into the recognition procedure.  The main point of the determination of  the recognition result verification module depends on the confidence rate of the transmitting result of the face recognition software.  In the worst case, when the identification fails, the times of retransmitting request is controlled by the greeting application according to a request time out, to determine whether the request has to be retransmitted.  If the correct result is not obtained after the identification is time out, the identification is failed.  Then the greeting application still can sends out a general greeting without identify information, such as a simple “Hello”.
Therefore, as the invention adopts the above technical solution, the invention does not need fixed the face pose to identify a face.
Meanwhile, the invention can identify a face based on a local or web server database, and it increases the accuracy of the identification.
Furthermore, the face recognition procedure meets the real‐time requirement through the suitable camera, the operation hardware module and the operation frame.
While the present disclosure has been described in connection with certain exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, the device and structure, which are not specifically described, should be understood as the common manner in the art to be implemented; any people skilled in the art can make possible changes and modifications, or equivalents thereof for the technical solution of the invention according to the above method without falling out of the scope of the invention.
Therefore, the various modifications and equivalent arrangements without departing away from the technical solution of the invention, are included within the spirit and the scope of the technical solution of the invention.

Claims (9)

What is claimed is:
1. An image recognition system, comprising: a robot, having an image collection module, configured to drive the image collection module to collect an image in a view of the robot by an image collection drive module; a light source configured to supplement light when the image collection module collects an image; a face detection module configured to locate a face image in the image collected by the image collection module; a face recognition module configured to implement a preprocessing for the located face image, then the preprocessed face image being compared with image feature information of a known identity contained in a database, to determine an identity information and a confidence rate of the located face image; a greeting application of the robot sends a recognition request to the image collection drive module, the image collection drive module accepts the recognition request, and drives the image collection module to collect the image and transmits the image to the face detection module; the face recognition module transmits the identity information and the confidence rate to result verification module, the system executes a corresponding operation according to a determining result as generated by the verification module; if the determining result determines the identity information is correct, the determining result is sent to the greeting application, and the greeting application utilizes a sounding device to greet user in the image of the image collection module according to a recognized identity; if the determining result determines the identity information is wrong, the system retransmits the recognition request to the image collection drive module, and reenters into a recognition procedure, the result verification module depends on the confidence rate to output the determining result.
2. The image recognition system as claimed in Claim 1, wherein the image collection module is a high definition camera, the high definition camera is capable to obtain static image and to collect at least 30 images per second; the high definition camera is connected to the robot by a MIPI or a USB interface.
3. The image recognition system as claimed in Claim 1, wherein the light source comprises an ambient light source and an infrared light source; upon utilizing the ambient light to supplement the light and collecting the image by the image collection module, if the collected image can not be recognized, the light is supplemented by the infrared light source.
4. The image recognition system as claimed in Claim 1, the preprocessing comprising: executing an angle correction treatment and a light treatment for the face image presenting in the image.
5. The image recognition system as claimed in Claim 1, wherein the database comprises a local data memory module and a web server data memory module.
6. The image recognition system as claimed in Claim 1, wherein the robot further comprises a sounding device connected to the database, the sounding device sends out various types of prompt tones according to a comparative result from the face recognition module.
7. The image recognition system as claimed in Claim 1, further comprising a record feedback device to record and/or feed back a comparative result from the face recognition module.
8. The image recognition system as claimed in Claim 1, wherein the face recognition module implements a comparison by SVM algorithm.
9. An image recognition method, using an image recognition system as claimed in any one of Claims 1‐8, comprising the steps of: (a) collecting an image in the view of the robot by the image collection module of the robot while supplementing the light by a light source; (b) using the face detection module to implement a location process to the face image presenting in the image which is collected by the image collection module; (c) using the face detection module to implement a preprocessing for the face image after being located, then comparing with an image feature information of a known identity contained in a database, to determine an identity information and a confidence rate of the located face image.
NZ734556A 2015-01-12 2015-06-12 Image recognition system and method NZ734556B2 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
CN201510014262.8A CN105844202A (en) 2015-01-12 2015-01-12 Image recognition system and method
CN201510014262.8 2015-01-12
PCT/CN2015/081403 WO2016112630A1 (en) 2015-01-12 2015-06-12 Image recognition system and method

Publications (2)

Publication Number Publication Date
NZ734556A NZ734556A (en) 2020-09-25
NZ734556B2 true NZ734556B2 (en) 2021-01-06

Family

ID=

Similar Documents

Publication Publication Date Title
CA2973393C (en) A system and a method for image recognition
Sugano et al. Appearance-based gaze estimation using visual saliency
US11055569B2 (en) Systems and methods for detecting a point of interest change using a convolutional neural network
US20180276465A1 (en) Method of recognition based on iris recognition and electronic device supporting the same
WO2015165365A1 (en) Facial recognition method and system
CN105759971A (en) Method and system for automatically prompting distance from human eyes to screen
US10685263B2 (en) System and method for object labeling
CN104166835A (en) Method and device for identifying living user
US9892315B2 (en) Systems and methods for detection of behavior correlated with outside distractions in examinations
US10741175B2 (en) Systems and methods for natural language understanding using sensor input
CN104143086A (en) Application technology of portrait comparison to mobile terminal operating system
WO2021197369A1 (en) Liveness detection method and apparatus, electronic device, and computer readable storage medium
US20180307896A1 (en) Facial detection device, facial detection system provided with same, and facial detection method
US11972634B2 (en) Image processing method and apparatus
Alburaiki et al. Mobile based attendance system: face recognition and location detection using machine learning
CN112949467B (en) Face detection method, device, electronic equipment and storage medium
AU2017101096A4 (en) A system and a method for image recognition
NZ734556B2 (en) Image recognition system and method
CN116311650A (en) Man-machine interaction method based on intelligent entrance guard's sight tracking and gesture recognition
CN112307966B (en) Event display method and device, storage medium and electronic equipment
OA18310A (en) A system and a method for image recognition.
CN113901934A (en) Intelligent visual detection method, system and device for large infusion package product
JP6801577B2 (en) Authentication system, authentication method, and gate
KR20210157052A (en) Object recognition method and object recognition apparatus
CN112529895B (en) Method, apparatus, device and storage medium for processing image