阅读说明：本技术 基于近红外光谱成像的人皮面具识别方法及系统 (Human skin mask identification method and system based on near infrared spectrum imaging ) 是由 刘仓理 张大勇 黄立贤 储松南 骆永全 沈志学 王海峰 曹宁翔 曾建成 于 2020-08-05 设计创作，主要内容包括：本发明公开了基于近红外光谱成像的人皮面具识别方法及系统,涉及光谱成像技术,其技术方案要点是：采集面部皮肤、人皮面具的近红外光谱数据,并根据近红外光谱数据建立光谱数据库；将光谱数据库中的近红外光谱数据输入光谱分类算法模型进行训练,得到光谱识别模型；通过近红外光谱成像系统对目标物成像,根据面部检测算法识别提取成像图像中的面部区域；提取近红外光谱序列图像中属于面部区域所有像素的面部光谱数据；将面部光谱数据输入光谱识别模型进行识别。基于人皮面具和人面部皮肤在1400-1500nm和1900-2000nm的光谱差异,结合光谱数据库将人皮面具识别出来,可适用于中远距离的识别,和图像识别相比,具有可远程、搜索范围宽、准确率高等优点。(The invention discloses a human skin mask identification method and a system based on near infrared spectrum imaging, which relate to the spectrum imaging technology and have the technical scheme key points that: collecting near infrared spectrum data of facial skin and human skin, and establishing a spectrum database according to the near infrared spectrum data; inputting the near infrared spectrum data in the spectrum database into a spectrum classification algorithm model for training to obtain a spectrum identification model; imaging the target object through a near infrared spectrum imaging system, and identifying and extracting a face area in an imaging image according to a face detection algorithm; extracting facial spectrum data of all pixels belonging to a facial area in the near infrared spectrum sequence image; and inputting the facial spectrum data into a spectrum recognition model for recognition. Based on the spectrum difference between the skin of the human skin and the skin of the human face at 1400-1500nm and 1900-2000nm, the skin of the human is identified by combining the spectrum database, so that the method can be suitable for medium-distance and long-distance identification, and has the advantages of long distance, wide search range, high accuracy and the like compared with image identification.)

1. The human skin mask identification method based on near infrared spectrum imaging is characterized by comprising the following steps:

collecting near infrared spectrum data of facial skin and human skin, and establishing a spectrum database according to the near infrared spectrum data;

inputting the near infrared spectrum data in the spectrum database into a spectrum classification algorithm model for training to obtain a spectrum identification model;

imaging the target object through a near infrared spectrum imaging system, and identifying and extracting a face area in an imaging image according to a face detection algorithm;

extracting facial spectrum data of all pixels belonging to a facial area in the near infrared spectrum sequence image;

and inputting the facial spectrum data into a spectrum recognition model for recognition, and outputting a recognition result.

2. The method for identifying a human skin mask based on near infrared spectral imaging as claimed in claim 1, wherein the spectral wavelength of the near infrared spectral imaging system is 1400-1500 nm.

3. The method for identifying a human skin mask based on near infrared spectral imaging as claimed in claim 1, wherein the spectral wavelength of the near infrared spectral imaging system is 1900-2000 nm.

4. The method for identifying a human skin mask based on near infrared spectral imaging as claimed in claim 1, wherein the wavelength of the near infrared spectral data in the spectral database is 900-2000 nm.

5. The method of claim 1, wherein the near infrared spectral imaging system is any one of an AOTF-based spectral imaging system, a snapshot-based spectral imaging system, and a Fourier spectral imaging system.

6. Human skin recognition system based on near infrared spectral imaging, characterized by includes:

the spectral database (101) is used for storing near infrared spectral data of facial skin and human skin;

the model construction module (102) is used for inputting the near infrared spectrum data in the spectrum database (101) into a spectrum classification algorithm model for training to obtain a spectrum identification model;

the imaging module (103) is used for imaging the target object through the near infrared spectrum imaging system and identifying and extracting a face area in the imaged image according to a face detection algorithm;

the spectrum extraction module (104) is used for extracting facial spectrum data of all pixels belonging to a facial area in the near infrared spectrum sequence image;

and the recognition module (105) is used for inputting the facial spectrum data into the spectrum recognition model for recognition and outputting a recognition result.

7. A computer terminal comprising a memory, a processor and a computer program stored in the memory and operable to execute the method of any of claims 1 to 5 on the processor.

8. A computer-readable medium, in which a computer program is stored which, when being processed and executed, carries out the steps of the method according to any one of claims 1 to 5.

Technical Field

The invention relates to a spectral imaging technology, in particular to a human skin mask identification method and system based on near infrared spectral imaging.

Background

The skin mask is also called as an easy mask, is made of silica gel materials, has the skin texture and the skin color of real people, is suitable for people with various skin colors and facial shapes to wear, and can achieve the effect of being fake and unreal to a certain extent after being worn. The human skin mask can be used for shooting movies and TV shows and being worn by patients with face burn. However, there are few lawbreakers who wear their masks to commit crimes. In 2010, a white adult of American nationality purchased a mask of a black person to rob a bank in the young, which is the first case of crime with a "face mask" known in the United states. Police without any related experience are puzzled, the detection direction is always locked as a black person, and finally, a girl of the young stands for a horse to alarm after finding the mask, so that the case is detected. Since 2013 in 2 months, many police in Guangzhou, Jiangsu and the like capture criminals who use 'human skins' to perform 'easy' and enter the room or specially aim at ATM to implement crimes. The police shows that the criminal can deceive the camera installed in the public place after wearing the skin mask, the police is difficult to master the true looks of the criminal, which brings higher difficulty to the case investigation, and the criminal will be more increased and has no fear.

At present, for the identification of a human skin mask, the identification can be carried out from the aspects of facial details, expressions, speaking voices and the like in the short-distance and long-time observation process. It has been reported that with iphone phones it is possible to identify real persons and human facial masks customized to the real person's face, mainly by virtue of the high pixel and detail resolution capabilities of the iphone phone camera. However, for the medium-distance and long-distance cameras, it is difficult to identify the human skin mask because sufficient human face details cannot be obtained.

The spectral imaging technology can divide the substance spectrum into dozens of to hundreds of wave bands by utilizing the optical filtering technology, so that the identification capability of the spectral imaging technology on the fine spectral characteristics of the target is greatly improved, and the spectral imaging technology has great application value in the aspects of camouflage identification, ground feature classification, environment monitoring and evaluation, resource remote sensing investigation, biological epidermis and organ inner wall surface health state diagnosis, monitoring of the physiological process of living tissues and the like. At present, no public report is found on a human skin mask recognition technology based on a spectral imaging analysis technology. Therefore, how to research and design a human skin mask identification method and system based on near infrared spectrum imaging is a problem which is urgently needed to be solved at present.

Disclosure of Invention

The invention aims to provide a method and a system for identifying a human skin mask based on near infrared spectrum imaging, which are based on spectral differences between the human skin mask and human face skin, are combined with a spectral database to identify the human skin mask, can be suitable for medium-distance and long-distance identification, have the advantages of long distance, wide search range, high accuracy and the like compared with image identification, can be suitable for important places, and realize real-time identification and monitoring of personnel wearing the human skin mask.

The technical purpose of the invention is realized by the following technical scheme:

in a first aspect, a human skin mask identification method based on near infrared spectral imaging is provided, which comprises the following steps:

collecting near infrared spectrum data of facial skin and human skin, and establishing a spectrum database according to the near infrared spectrum data;

inputting the near infrared spectrum data in the spectrum database into a spectrum classification algorithm model for training to obtain a spectrum identification model;

imaging the target object through a near infrared spectrum imaging system, and identifying and extracting a face area in an imaging image according to a face detection algorithm;

extracting facial spectrum data of all pixels belonging to a facial area in the near infrared spectrum sequence image;

and inputting the facial spectrum data into a spectrum recognition model for recognition, and outputting a recognition result.

Preferably, the spectral wavelength of the near infrared spectrum imaging system is 1400-1500 nm.

Preferably, the spectral wavelength of the near infrared spectrum imaging system is 1900-2000 nm.

Preferably, the wavelength of the near infrared spectrum data in the spectrum database is 900-2000 nm.

Preferably, the near infrared spectral imaging system is any one of an AOTF-based spectral imaging system, a snapshot spectral imaging system and a fourier spectral imaging system.

In a second aspect, there is provided a human skin mask identification system based on near infrared spectral imaging, comprising:

the spectral database is used for storing near infrared spectral data of facial skin and human skin;

the model construction module is used for inputting the near infrared spectrum data in the spectrum database into a spectrum classification algorithm model for training to obtain a spectrum identification model;

the imaging module is used for imaging the target object through the near infrared spectrum imaging system and identifying and extracting a face area in the imaged image according to a face detection algorithm;

the spectrum extraction module is used for extracting facial spectrum data of all pixels belonging to a facial area in the near infrared spectrum sequence image;

and the recognition module is used for inputting the facial spectrum data into the spectrum recognition model for recognition and outputting a recognition result.

In a third aspect, there is provided a computer terminal comprising a memory, a processor and a computer program stored in the memory and executable in the processor to perform the method of any one of the first aspects.

In a fourth aspect, there is provided a computer readable medium storing a computer program which when processed and executed performs the steps of the method according to any one of the first aspect.

Compared with the prior art, the invention has the following beneficial effects: the invention is based on the spectral difference between the skin and the facial skin, especially in the wavelength ranges of 1400-1500nm and 1900-2000nm, and the skin is identified by combining the spectral database, so that the invention can be suitable for medium and long distance identification, and compared with image identification, the invention has the advantages of long distance, wide search range, high accuracy and the like; the invention provides a human skin mask recognition system which can be installed in important places to realize real-time recognition and monitoring of people wearing the human skin mask.

Drawings

The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:

FIG. 1 is a flowchart of the operation in embodiment 1 of the present invention;

FIG. 2 is a functional block diagram in embodiment 2 of the present invention;

FIG. 3 is a graph of the reflectance spectrum of facial skin and a human skin mask in an embodiment of the present invention.

Reference numbers and corresponding part names in the drawings:

101. a spectral database; 102. a model building module; 103. an imaging module; 104. a spectrum extraction module; 105. and identifying the module.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail in the following with reference to examples 1-2 and accompanying drawings 1-3, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not to be construed as limiting the present invention.