阅读说明：本技术 一种基于3d结构光的人脸识别系统及方法 (Face recognition system and method based on 3D structured light ) 是由 莫君 李竹林 于 2021-08-31 设计创作，主要内容包括：本发明公开了一种基于3D结构光的人脸识别系统及方法,包括：终端设备,用于对用户进行身份注册及密态人脸特征采集,并基于密态人脸特征生成特征码,将特征码和密态人脸特征发送至云服务器和进口验证端；进口验证端,接收终端设备所发送的特征码进行存储；云服务器,接收终端设备所发送的特征码和密态人脸特征进行存储,并向商品验证端发送与该特征码相匹配的密态人脸特征；商品验证端,接收云服务器所发送的密态人脸特征进行存储,并将正/负价格信息与匹配的密态人脸特征发送至出口验证端；出口验证端,接收商品验证端发送的正/负价格信息及与其匹配的密态人脸特征,并进行结算；该基于3D结构光的人脸识别系统及方法能够避免在结算时出现堵塞。(The invention discloses a face recognition system and method based on 3D structured light, which comprises the following steps: the terminal equipment is used for carrying out identity registration and secret face feature acquisition on the user, generating a feature code based on the secret face feature, and sending the feature code and the secret face feature to the cloud server and the import verification end; the import verification end receives and stores the feature codes sent by the terminal equipment; the cloud server receives the feature codes and the secret face features sent by the terminal equipment, stores the feature codes and the secret face features, and sends the secret face features matched with the feature codes to the commodity verification end; the commodity verification end receives and stores the secret-state face features sent by the cloud server, and sends the positive/negative price information and the matched secret-state face features to the export verification end; the export verification end receives the positive/negative price information sent by the commodity verification end and the secret face characteristics matched with the positive/negative price information and settles accounts; the face recognition system and method based on the 3D structured light can avoid blockage during settlement.)

1. A face recognition system based on 3D structured light, comprising:

the terminal equipment is used for carrying out identity registration and secret face feature acquisition on the user, generating a feature code based on the secret face feature, and sending the feature code and the secret face feature to the cloud server and the import verification end;

the import verification end receives and stores the feature codes sent by the terminal equipment, and sends the feature codes to the cloud server, and the import verification end comprises an input unit and an import gate unit;

the cloud server receives the feature codes and the secret face features sent by the terminal equipment, stores the feature codes and the secret face features, and sends the secret face features matched with the feature codes to the commodity verification end;

the commodity verification end receives and stores the secret-state face features sent by the cloud server, and sends the positive/negative price information and the matched secret-state face features to the export verification end, wherein the commodity verification end comprises a track unit, a first face identification unit and a price unit;

and the export verification end receives the positive/negative price information sent by the commodity verification end and the secret face characteristics matched with the positive/negative price information and settles the account, and comprises a settlement unit, a second face recognition unit and an export gate unit.

2. The 3D structured light based face recognition system of claim 1, wherein: the cloud server also comprises a matching unit which is used for searching the secret face features corresponding to the feature codes from the secret face features of a plurality of users stored in the cloud server according to the feature codes of the users for matching.

3. The 3D structured light based face recognition system of claim 1, wherein: the system also comprises a deleting unit which is used for deleting the secret-state face features of the commodity verifying end and the export verifying end, and deleting the corresponding secret-state face features in the commodity verifying end and the export verifying end after the export verifying end completes settlement.

4. The 3D structured light based face recognition system of claim 1, wherein: the terminal equipment comprises a mobile intelligent terminal and a fixed intelligent terminal, and the mobile intelligent terminal and the fixed intelligent terminal are man-machine interaction equipment with image acquisition, data acquisition and data transmission functions.

5. The 3D structured light based face recognition system of claim 1, wherein: the first face recognition unit and the second face recognition unit acquire the secret-state face features by combining a camera with an infrared structural optical transceiver.

6. The 3D structured light based face recognition system of claim 1, wherein: the input unit comprises a feature code reading unit and a feature code input unit, the feature code reading unit reads the feature code on the feature code verification card through NFC communication, and the feature code input unit directly inputs the feature code through a touch screen.

7. A face recognition method based on 3D structured light is characterized by comprising the following steps:

the terminal equipment registers the identity of the user and acquires the face features of the secret state, and generates feature codes based on the face features of the secret state;

matching the secret face features with the feature codes and then sending the matched features to a cloud server and an import verification end;

the import verification end matches the generated feature code according to the feature code input by the user and judges whether to open the import gate unit or not;

when the gate is judged to be opened, the cloud server sends the secret-state human face features matched with the input feature codes to the commodity verification end;

the commodity verification end collects the face characteristics of the user at a commodity shelf, and when the collection is successful, the shelf is opened, and the commodity taking-out or putting-in track is judged;

sending corresponding positive/negative price information to an export verification end according to the taking-out or putting-in track of the commodity;

the exit verification end collects the face characteristics of the user, and when the collection is successful, positive/negative price information under the secret face characteristic name is matched, and statistical settlement is carried out;

after settlement is successful, the exit gate unit is opened.

8. The 3D structured light based face recognition method according to claim 7, wherein: and after the exit gate unit is opened, the deleting unit sends a deleting command to delete the secret human face features stored in the commodity verifying end and the exit verifying end.

9. The face recognition method based on 3D structured light as claimed in claim 7, wherein the method for determining the commodity taking out or putting in trajectory is:

collecting video image information at a side view angle, dividing the video image information into nine-square lattices based on the video image information, and assigning a value of zero to each area;

when a goods shelf is opened, scanning is carried out, when a nine-square grid area obtains a commodity sample, the value of the area is increased by one, and when the nine-square grid area loses the commodity sample, the value of the area is reset to zero;

and judging whether the commodity sample is a taking out/putting in action based on the assignment change trend of the Sudoku area.

10. The 3D structured light based face recognition method according to any one of claims 7 to 9, being used in an unmanned vending store.

Technical Field

The invention relates to a face recognition system and method based on 3D structured light.

Background

The structured light is a set of system structures consisting of a projector and a camera. The projector is used for projecting specific light information to the surface of an object and the background, and the specific light information is collected by the camera. Information such as the position and depth of the object is calculated from the change of the optical signal caused by the object, and the entire three-dimensional space is restored.

The basic principle of the 3D structured light technology is that light rays with certain structural characteristics are projected to a shot object through a near-infrared laser, and then collected through a special infrared camera. The light with a certain structure can acquire different image phase information according to different depth areas of a shot object, and then the change of the structure is converted into depth information through an arithmetic unit, so that a three-dimensional structure is obtained. In short, the three-dimensional structure of the object to be photographed is acquired by an optical means, and the acquired information is applied more deeply.

Based on this, 3D structured light is often used in practical application scenes such as security, payment, and access control at present, and in order to ensure the security of user data, a face recognition system is usually used to identify a user object in a current scene.

At present, some face recognition systems are already applied to markets, but the face recognition systems are generally applied to self-service checkout vending machines, and users select face recognition to settle accounts after scanning barcodes of various commodities by themselves.

Disclosure of Invention

Aiming at the defects in the prior art, the invention aims to provide a face recognition system and a face recognition method based on 3D structured light, which can avoid the occurrence of blockage during settlement.

The technical scheme adopted by the invention for solving the technical problems is as follows:

a 3D structured light based face recognition system comprising:

the terminal equipment is used for carrying out identity registration and secret face feature acquisition on the user, generating a feature code based on the secret face feature, and sending the feature code and the secret face feature to the cloud server and the import verification end;

the import verification end receives and stores the feature codes sent by the terminal equipment, and sends the feature codes to the cloud server, and the import verification end comprises an input unit and an import gate unit;

the cloud server receives the feature codes and the secret face features sent by the terminal equipment, stores the feature codes and the secret face features, and sends the secret face features matched with the feature codes to the commodity verification end;

the commodity verification end receives and stores the secret-state face features sent by the cloud server, and sends the positive/negative price information and the matched secret-state face features to the export verification end, wherein the commodity verification end comprises a track unit, a first face identification unit and a price unit;

and the export verification end receives the positive/negative price information sent by the commodity verification end and the secret face characteristics matched with the positive/negative price information and settles the account, and comprises a settlement unit, a second face recognition unit and an export gate unit.

Preferably, the cloud server further comprises a matching unit, configured to search for a dense-state face feature corresponding to the feature code from among the dense-state face features of the users stored in the cloud server according to the feature code of the user, and perform matching.

Preferably, the system further comprises a deleting unit, configured to delete the secret-state face features according to the commodity verification end and the export verification end, and delete the corresponding secret-state face features in the commodity verification end and the export verification end after the export verification end completes settlement.

Preferably, the terminal device comprises a mobile intelligent terminal and a fixed intelligent terminal, and the mobile intelligent terminal and the fixed intelligent terminal are human-computer interaction devices with image acquisition, data acquisition and data transmission functions.

Preferably, the first face recognition unit and the second face recognition unit acquire the dense face features by combining a camera with an infrared structural optical transceiver.

Preferably, the input unit comprises a feature code reading unit and a feature code input unit, the feature code reading unit reads the feature code on the feature code verification card through NFC communication, and the feature code input unit directly inputs the feature code through a touch screen.

Another technical problem to be solved by the present invention is to provide a face recognition method based on 3D structured light, comprising:

the terminal equipment registers the identity of the user and acquires the face features of the secret state, and generates feature codes based on the face features of the secret state;

matching the secret face features with the feature codes and then sending the matched features to a cloud server and an import verification end;

the import verification end matches the generated feature code according to the feature code input by the user and judges whether to open the import gate unit or not;

when the gate is judged to be opened, the cloud server sends the secret-state human face features matched with the input feature codes to the commodity verification end;

the commodity verification end collects the face characteristics of the user at a commodity shelf, and when the collection is successful, the shelf is opened, and the commodity taking-out or putting-in track is judged;

sending corresponding positive/negative price information to an export verification end according to the taking-out or putting-in track of the commodity;

the exit verification end collects the face characteristics of the user, and when the collection is successful, positive/negative price information under the secret face characteristic name is matched, and statistical settlement is carried out;

after settlement is successful, the exit gate unit is opened.

Preferably, after the exit gate unit is opened, the deleting unit sends a deleting command to delete the secret-state human face features stored in the commodity verification end and the exit verification end.

Preferably, the method for judging the commodity taking out or putting in trajectory is as follows:

collecting video image information at a side view angle, dividing the video image information into nine-square lattices based on the video image information, and assigning a value of zero to each area;

when a goods shelf is opened, scanning is carried out, when a nine-square grid area obtains a commodity sample, the value of the area is increased by one, and when the nine-square grid area loses the commodity sample, the value of the area is reset to zero;

and judging whether the commodity sample is a taking out/putting in action based on the assignment change trend of the Sudoku area.

Preferably, the system is used in the vending market.

The invention has the beneficial effects that:

compare in traditional market and carry out settlement one by one in the final stage, this scheme has adopted the mode of precomputation, via carrying out the collection of commodity positive/negative price in the goods shelves stage to send the collection result to the exit end, finally count up via the exit end, avoided traditional market to block up the defect of settling accounts in the exit end stage, even when the passenger flow volume is great, settlement that also can be quick, simultaneously, this technique also is more convenient for the popularization and use of unmanned selling goods.

Detailed Description

The present invention is further described with reference to specific examples to enable those skilled in the art to better understand the present invention and to practice the same, but the examples are not intended to limit the present invention.

Examples

A 3D structured light based face recognition system comprising:

the terminal equipment is used for carrying out identity registration and secret face feature acquisition on the user, generating a feature code based on the secret face feature, and sending the feature code and the secret face feature to the cloud server and the import verification end;

the import verification end receives and stores the feature codes sent by the terminal equipment, and sends the feature codes to the cloud server, and the import verification end comprises an input unit and an import gate unit;

the cloud server receives the feature codes and the secret face features sent by the terminal equipment, stores the feature codes and the secret face features, and sends the secret face features matched with the feature codes to the commodity verification end;

the commodity verification end receives and stores the secret-state face features sent by the cloud server, and sends the positive/negative price information and the matched secret-state face features to the export verification end, wherein the commodity verification end comprises a track unit, a first face identification unit and a price unit;

and the export verification end receives the positive/negative price information sent by the commodity verification end and the secret face characteristics matched with the positive/negative price information and settles the account, and comprises a settlement unit, a second face recognition unit and an export gate unit.

The cloud server also comprises a matching unit which is used for searching the secret face features corresponding to the feature codes from the secret face features of a plurality of users stored in the cloud server according to the feature codes of the users for matching.

The system also comprises a deleting unit which is used for deleting the secret-state face features of the commodity verifying end and the export verifying end, and deleting the corresponding secret-state face features in the commodity verifying end and the export verifying end after the export verifying end completes settlement.

The terminal equipment comprises a mobile intelligent terminal and a fixed intelligent terminal, and the mobile intelligent terminal and the fixed intelligent terminal are man-machine interaction equipment with image acquisition, data acquisition and data transmission functions.

The first face recognition unit and the second face recognition unit acquire the secret-state face features by combining a camera with an infrared structural optical transceiver.

The input unit comprises a feature code reading unit and a feature code input unit, the feature code reading unit reads the feature code on the feature code verification card through NFC communication, and the feature code input unit directly inputs the feature code through a touch screen.

A face recognition method based on 3D structured light comprises the following steps:

the terminal equipment registers the identity of the user and acquires the face features of the secret state, and generates feature codes based on the face features of the secret state;

matching the secret face features with the feature codes and then sending the matched features to a cloud server and an import verification end;

the import verification end matches the generated feature code according to the feature code input by the user and judges whether to open the import gate unit or not;

when the gate is judged to be opened, the cloud server sends the secret-state human face features matched with the input feature codes to the commodity verification end;

the commodity verification end collects the face characteristics of the user at a commodity shelf, and when the collection is successful, the shelf is opened, and the commodity taking-out or putting-in track is judged;

sending corresponding positive/negative price information to an export verification end according to the taking-out or putting-in track of the commodity;

the exit verification end collects the face characteristics of the user, and when the collection is successful, positive/negative price information under the secret face characteristic name is matched, and statistical settlement is carried out;

after settlement is successful, the exit gate unit is opened.

And after the exit gate unit is opened, the deleting unit sends a deleting command to delete the secret human face features stored in the commodity verifying end and the exit verifying end.

The method for judging the commodity taking-out or putting-in track comprises the following steps:

collecting video image information at a side view angle, dividing the video image information into nine-square lattices based on the video image information, and assigning a value of zero to each area;

when a goods shelf is opened, scanning is carried out, when a nine-square grid area obtains a commodity sample, the value of the area is increased by one, and when the nine-square grid area loses the commodity sample, the value of the area is reset to zero;

and judging whether the commodity sample is a taking out/putting in action based on the assignment change trend of the Sudoku area.

Based on the characteristic of commodity taking-out or putting-in track judgment, in the scheme, the commodity verification end is arranged at the side position of the goods shelf, commodities on the goods shelf are arranged in layers, each layer corresponds to one track unit, meanwhile, commodities on the goods shelf are guaranteed in a box, a commodity identification feature is labeled on one side corresponding to the track unit, after the commodities are settled, the cloud server sends a command to enable the commodity identification feature to be deleted from the track unit, and when new commodities are put in a warehouse, the cloud server sends a command to enable the commodity identification unit to be re-recorded into a storage chip in the track unit.

The scheme can be used for the places of the unmanned selling markets, particularly the entrances and exits of the unmanned selling markets and the goods shelves.

In this scheme, also can adopt TOF technique to replace 3D structured light technique and carry out face identification, nevertheless because the factor of applied scene adopts 3D structured light technique more to be applicable to this kind of static scene environment partially of unmanned shopping mall.

The invention has the beneficial effects that:

compare in traditional market and carry out settlement one by one in the final stage, this scheme has adopted the mode of precomputation, via carrying out the collection of commodity positive/negative price in the goods shelves stage to send the collection result to the exit end, finally count up via the exit end, avoided traditional market to block up the defect of settling accounts in the exit end stage, even when the passenger flow volume is great, settlement that also can be quick, simultaneously, this technique also is more convenient for the popularization and use of unmanned selling goods.

The above-described embodiments of the present invention are not intended to limit the scope of the present invention, and the embodiments of the present invention are not limited thereto, and various other modifications, substitutions and alterations can be made to the above-described structure of the present invention without departing from the basic technical concept of the present invention as described above, according to the common technical knowledge and conventional means in the field of the present invention.