CN114463792A - Multispectral identification method, multispectral identification device, multispectral identification equipment and readable storage medium - Google Patents
Multispectral identification method, multispectral identification device, multispectral identification equipment and readable storage medium Download PDFInfo
- Publication number
- CN114463792A CN114463792A CN202210125616.6A CN202210125616A CN114463792A CN 114463792 A CN114463792 A CN 114463792A CN 202210125616 A CN202210125616 A CN 202210125616A CN 114463792 A CN114463792 A CN 114463792A
- Authority
- CN
- China
- Prior art keywords
- spectral
- spectrum
- image
- target
- images
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Image Processing (AREA)
Abstract
The application discloses a multispectral identification method, a multispectral identification device, multispectral identification equipment and a readable storage medium, wherein the multispectral identification method comprises the following steps: the double-buffer collector is used for collecting a plurality of spectral images to be detected, after one buffer collector finishes collection, the collected spectral images can be spliced and input into the spectral detection model for detection, and when the spectral images are detected, the other buffer collection image starts to collect the spectral images, so that the spectral image collection and the spectral image detection are carried out simultaneously, and the spectral image collection efficiency can be improved. Moreover, the plurality of spectral images are spliced and detected, so that each spectral image can be prevented from being detected, the detection times are reduced, the detection results of the plurality of spectral images can be determined only according to the splicing mode, the time for detecting the spectral images can be effectively shortened, and the spectral detection efficiency is improved.
Description
Technical Field
The present application relates to the field of biometric identification technologies, and more particularly, to a multispectral identification method, apparatus, device, and readable storage medium.
Background
In recent years, it has become common to verify identity by using biometric identification technology, and spectral identification technology is one of the technologies applied to biometric identification, and can be applied to devices for identity authentication, such as attendance machines, door controllers, and gateway gates.
The existing multispectral object identification method generally obtains images of each spectrum independently, detects the images of each spectrum one by one, and then integrates the detection results of all the images of the spectrum. However, in the existing spectrum identification method, under the condition that a large number of spectrum images need to be detected, if the performance of the inspection equipment is insufficient, the detection process is not smooth and even stuck, and the spectrum identification efficiency is further influenced.
Therefore, how to improve the efficiency of spectrum identification is a problem worthy of study.
Disclosure of Invention
In view of the above, the present application provides a multispectral identification method, device, apparatus and readable storage medium, which are used to improve the efficiency of spectral identification.
In order to achieve the above object, the following solutions are proposed:
a method of multispectral identification, comprising:
collecting spectrum images of a plurality of organisms to be detected by using a double-buffer collector, wherein two buffer collectors of the double-buffer collector are used for collecting the spectrum images in sequence, and the spectrum image collected by each buffer collector is a spectrum image obtained by splicing the set number of spectrum images according to a set splicing mode;
when the target spectrum images collected by one target buffer collector in the double buffer collectors reach the set number, inputting the spliced spectrum images into a spectrum detection model to obtain a detection result of the spliced spectrum images, wherein the spectrum detection model is used for detecting the spectrum contained in the spectrum images;
determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral image;
and for each target spectral image, performing spectral recognition on the target spectral image according to a detection result corresponding to the target spectral image to obtain a spectral recognition result of the target spectral image.
Preferably, the process of stitching the spectral images includes:
for each buffer collector:
establishing a storage space according to the set number of spectral images, wherein the storage space is provided with a plurality of storage positions;
and splicing each spectral image to a corresponding storage position in the storage space according to a set splicing mode to obtain a spliced spectral image.
Preferably, before performing spectrum recognition on the target spectrum image, the method further comprises:
for each target spectral image, inputting a detection result corresponding to the target spectral image into an anti-counterfeiting identification model to obtain a true and false judgment result of the target spectral image;
the performing spectrum identification on the target spectrum image comprises:
and carrying out spectrum recognition on the target spectrum image with the result characterized as a real organism to obtain a spectrum recognition result of the target spectrum image.
Preferably, the determining, according to the set stitching method and the detection result of the stitched spectral image, the detection result corresponding to each target spectral image includes:
determining the region of each target spectral image after splicing according to the set splicing mode;
and determining the detection result of each region of the spliced spectral image based on the detection result of the spliced spectral image, and taking the detection result of each region as the detection result of the target spectral image corresponding to the region.
Preferably, the performing, for each target spectral image, spectral recognition on the target spectral image according to a detection result corresponding to the target spectral image to obtain a spectral recognition result of the target spectral image includes:
for a target spectrum image with a detection result only having a single spectrum, performing spectrum identification on the single spectrum of the target spectrum image by using a target identification threshold higher than an initial identification threshold corresponding to the single spectrum to obtain a spectrum identification result of the target spectrum image reaching the target identification threshold;
and carrying out spectrum recognition on a plurality of spectrums of the target spectrum image with a plurality of spectrums of the detection result, and if the first two recognized recognition results are consistent, determining the first two recognized recognition results as the spectrum recognition results of the target spectrum image.
Preferably, the training process of the anti-counterfeiting recognition model comprises the following steps:
acquiring training spectral images of different prosthesis types and containing prosthesis characteristics;
inputting the training spectral image marked with the prosthesis type into the anti-counterfeiting recognition model to obtain a predicted prosthesis type output by the anti-counterfeiting recognition model;
and updating the parameters of the anti-counterfeiting recognition model by taking the output prosthesis type which is predicted to approach the prosthesis type marked by the training spectral image as a training target.
A multispectral identification device, comprising:
the image acquisition unit is used for acquiring spectral images of a plurality of organisms to be detected by using a double-buffer collector, two buffer collectors of the double-buffer collector are used for sequentially acquiring the spectral images, and the spectral images acquired by each buffer collector are the spectral images acquired by splicing the set number of spectral images according to a set splicing mode;
the image detection unit is used for inputting the spliced spectral images to a spectral detection model when the target spectral images acquired by one target buffer collector in the double buffer collectors reach the set number, so as to obtain the detection result of the spliced spectral images, and the spectral detection model is used for detecting the spectrum contained in the spectral images;
the detection result determining unit is used for determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral image;
and the image identification unit is used for carrying out spectrum identification on the target spectrum image according to the detection result corresponding to the target spectrum image aiming at each target spectrum image to obtain the spectrum identification result of the target spectrum image.
Preferably, the image acquisition unit includes:
the storage space establishing unit is used for establishing a storage space for each buffer collector according to the set number of spectral images, and the storage space is provided with a plurality of storage positions;
and the image splicing unit is used for splicing each spectral image to the corresponding storage position in the storage space according to a set splicing mode aiming at each buffer collector to obtain the spliced spectral image.
A multispectral identification device comprising a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program and realizing the steps of the multispectral identification method.
A readable storage medium having stored thereon a computer program which, when executed by a processor, carries out the steps of the above-mentioned multi-spectral recognition method.
According to the scheme, the multispectral identification method provided by the application comprises the following steps: the double-buffer collector is used for collecting a plurality of spectral images to be detected, after one buffer collector finishes collecting, the collected spectral images can be spliced and input into the spectral detection model for detection, and when the spectral images are detected, the other buffer collection image starts to collect the spectral images, so that the spectral images are collected and detected at the same time, and the efficiency of collecting the spectral images can be improved.
Moreover, the plurality of spectral images are spliced and detected, so that each spectral image can be prevented from being detected, the detection times are reduced, the detection results of the plurality of spectral images can be determined only according to the splicing mode, the time for detecting the spectral images can be effectively shortened, and the spectral detection efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a schematic flowchart of a multispectral identification method according to an embodiment of the present disclosure;
fig. 2 is a specific example diagram of a detection result of a spliced spectral image according to an embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a multispectral identification device according to an embodiment of the present disclosure;
fig. 4 is a block diagram of a hardware structure of a multispectral identification device disclosed in an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Referring to fig. 1, fig. 1 is a schematic flow chart of a multispectral identification method provided in an embodiment of the present application, where the multispectral identification method includes:
step S100: and collecting spectral images of a plurality of organisms to be detected by using a double-buffer collector.
Specifically, two buffer collectors of the double-buffer collector can be used for sequentially collecting spectral images, and the spectral images collected by each buffer collector can be spectral images obtained by splicing a set number of spectral images according to a set splicing mode.
The spectral image of the biological body to be detected may be a spectral image including the characteristics of the biological body to be authenticated, such as: a spectral image of a human palm, a spectral image of a human fingerprint, a spectral image of a human face, and the like.
In addition, after any one buffer collector finishes the collection of the spectral image, the other buffer collector can be immediately used for collecting the spectral image, and the collected spectral image can be used for the subsequent steps.
Step S110: and when the target spectrum images collected by one target buffer collector in the double buffer collectors reach the set number, inputting the spliced spectrum images into a spectrum detection model to obtain the detection result of the spliced spectrum images.
Specifically, any one of the double-buffer collectors that collects a set number of spectral images may be used as a target buffer collector, the spectral images collected by the target buffer collector may be used as target spectral images, and the spectral images obtained by splicing the set number of target spectral images may be input to the spectral detection model to obtain a detection result of the spliced spectral images output by the spectral detection model. The detection result may include a detection result corresponding to each target spectral image.
The spectral detection model described above can be used to detect the spectrum contained in the spectral image, and the spectral detection model herein can be implemented based on a variety of methods, such as: detection models implemented based on the framework centernet2, as well as spectral detection models implemented by alternative methods, can be used.
Step S120: and determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral images.
Specifically, the splicing mode may be determined according to actual conditions, and taking two spectral images with a set number as an example, the two spectral images may be spliced up and down, or left and right, or the two spectral images may be spliced after being adjusted to a certain extent.
And determining the detection result corresponding to each target spectral image according to the detection result of the spliced spectral image obtained in the step and the splicing mode of the target spectral image.
Step S130: and for each target spectral image, performing spectral recognition on the target spectral image according to a detection result corresponding to the target spectral image to obtain a spectral recognition result of the target spectral image.
Specifically, the detection result of the spliced spectral image obtained in the above step may include a detection result corresponding to each target spectral image, so that the corresponding target spectral image may be subjected to spectral recognition according to the detection result corresponding to each target spectral image, and a spectral recognition result of each target spectral image may be obtained.
The obtained spectrum recognition result can be used for identity authentication of the organism corresponding to the target spectrum image.
According to the scheme, after a set number of spectral images are collected by any one buffer collector, the spliced spectral images can be detected and identified, meanwhile, another buffer collector can be used for collecting the spectral images, the detection of the spliced spectral images and the collection of other spectral images are carried out simultaneously, the time difference can be effectively utilized, and therefore the efficiency is improved.
In some embodiments of the present application, the above step S100 is introduced, a process of acquiring spectral images of a plurality of organisms to be detected by using a double-buffer collector, and then obtaining a spliced spectral image is performed, and then, a process of splicing the spectral images will be further described.
Specifically, for each buffer collector, the process may include the following steps:
and S1, establishing a storage space according to the set number of spectral images, wherein the storage space is provided with a plurality of storage positions.
Specifically, a storage space may be established according to a set number of spectrograms, the storage space may have a plurality of storage locations, and each storage location may store a corresponding spectral image.
And S2, splicing each spectral image to a corresponding storage position in the storage space according to a set splicing mode to obtain a spliced spectral image.
Specifically, for each spectral image, the spectral image can be spliced to the corresponding storage position in the storage space according to the set splicing mode, and after the set number of spectral images are all spliced to the corresponding storage positions, the spliced spectral image can be obtained.
According to the scheme, the storage space is established for storing the spliced spectral images, the storage space can be prevented from being established for each spectral image, and therefore the process of integrating the spectral images and splicing the spectral images can be omitted.
In some embodiments of the present application, step S120 is introduced, and a detection result corresponding to each target spectral image is determined according to the set stitching manner and the detection result of the stitched spectral image.
Specifically, the process may include the steps of:
and S1, determining the region of each target spectral image after splicing according to the set splicing mode.
Specifically, the region of each target spectral image may be determined in the spliced spectral images, wherein the same target spectral image may correspond to different regions in different splicing manners.
And S2, determining the detection result of each region of the spliced spectral images based on the detection results of the spliced spectral images, and taking the detection result of each region as the detection result of the target spectral image corresponding to the region.
Specifically, the detection result of the spliced spectral image may include the detection result of each target spectral image, so that the detection result of the region corresponding to each target spectral image may be determined, and the detection result of the region may be used as the detection result of the target spectral image.
Next, a process of determining a detection result of each target spectral image according to the stitching manner will be described in detail with reference to fig. 2.
Specifically, the set splicing manner may be multiple, and here, the set number of spectral images is two spectral images, and the splicing manner of the two spectral images is left-right splicing.
And (4) establishing a plane coordinate system by taking the vertex of the upper left corner of the detection result of the spliced spectral image as a coordinate origin. Because the two target spectral images have the same size and are spliced left and right, the corresponding area of each target spectral image can be determined in the range of the abscissa.
Referring to fig. 2, a region having an abscissa range of [0, width) may be taken as the region of the first target spectral image, and a region having an abscissa range of [ width, 2width) may be taken as the region of the second target spectral image.
And then determining the detection results of the two regions in the detection results of the spliced spectral images based on the determined region range, wherein the detection result of each region can be used as the detection result of the target spectral image corresponding to the region.
According to the scheme, the spliced spectral images can be detected by combining multiple detections of multiple target spectral images into one time, so that the time required by the detection of the target spectral images is shortened.
In some embodiments of the present application, the above step S130 is introduced, and a process of performing spectrum recognition on each target spectrum image according to a detection result corresponding to the target spectrum image to obtain a spectrum recognition result of the target spectrum image is described, and the process will be further described below.
Specifically, different detection results may correspond to different processing steps.
S1, the target spectrum image in which only a single spectrum exists for the detection result:
and carrying out spectrum recognition on the single spectrum of the target spectrum image by using a target recognition threshold higher than the initial recognition threshold corresponding to the single spectrum to obtain a spectrum recognition result of the target spectrum image reaching the target recognition threshold.
Specifically, each spectrum may have a corresponding initial recognition threshold, and if a single spectrum in the detection result reaches the initial recognition threshold corresponding to a certain spectrum, the initial recognition threshold may be increased to a target recognition threshold, where the increased amplitude may be determined according to an actual situation, for example: raise the initial recognition threshold by 20%, 30%, or 40%, etc.
Then, the single spectrum may be identified by the target identification threshold, and if the identification result of the single spectrum reaches the target identification threshold, the single spectrum included in the target spectrum image may be considered as the spectrum corresponding to the target identification threshold.
S2, a target spectrum image in which a plurality of spectra exist for the detection result:
and performing spectrum recognition on a plurality of spectra of the target spectrum image, and if the first two recognized recognition results are consistent, determining the first two recognized recognition results as the spectrum recognition results of the target spectrum image.
Specifically, for the identification of a plurality of spectra included in the target spectral image, each spectrum may have a corresponding initial identification threshold, and if the initial identification threshold of a certain spectrum is reached, the spectrum may be considered as the spectrum corresponding to the initial identification threshold.
If the first two identified identification results are consistent, that is, the first two identified spectra are the same spectrum, the identification results of the first two identified spectra can be used as the identification results of the target spectral image.
It can be seen from the above solutions that, in the embodiments of the present application, different identification strategies are applied for detection results in which a single spectrum and multiple spectra exist, and thus, the case of variability can be flexibly dealt with.
In consideration of the fact that the acquired spectral images are not necessarily all spectral images of the organism to be authenticated, and may also be spectral images of the prosthesis, in the embodiment of the present application, before performing spectral recognition on the target spectral image, an anti-counterfeiting detection process for the spectral images may be added, so as to eliminate interference of the prosthesis.
Specifically, the process may include the steps of:
and inputting a detection result corresponding to each target spectral image into an anti-counterfeiting identification model to obtain a true and false judgment result of the target spectral image.
Specifically, the detection result corresponding to each target spectral image may be input to the anti-counterfeiting recognition model to obtain the predicted anti-counterfeiting score of the detection result of each target spectral image, and for the detection result reaching the set anti-counterfeiting score threshold, the target spectral image corresponding to the detection result may be used as the spectral image of the real organism, or else as the spectral image of the prosthesis.
The anti-counterfeiting identification model can be realized based on various methods, such as: the anti-counterfeiting identification model can be realized by using an anti-counterfeiting identification model based on the resnet18 and modified according to actual conditions, and other alternative methods.
Furthermore, one or more spectral anti-counterfeiting may be performed for the spectra contained in the target spectral image, such as: anti-counterfeiting can be performed for the visible light spectrum and/or for the near infrared spectrum, as well as other optional spectrum anti-counterfeiting.
After the anti-counterfeiting identification, performing spectrum identification on the target spectrum image, which may include:
and carrying out spectrum recognition on the target spectrum image with the result characterized as a real organism to obtain a spectrum recognition result of the target spectrum image.
Specifically, the process of performing spectrum recognition on the target spectrum image whose result is characterized by the real organism may refer to the process of performing spectrum recognition on the target spectrum image described above, and details are not repeated here.
In some embodiments of the present application, a process of inputting the detection result corresponding to the target spectral image to an anti-counterfeit recognition model to obtain a true and false determination result of the target spectral image is introduced, and a training process of the anti-counterfeit recognition model is introduced next.
Specifically, the process may include the steps of:
and S1, acquiring training spectral images of different prosthesis types and containing prosthesis characteristics.
In particular, for different types of prostheses, spectral images containing their features may be acquired as training spectral images.
And S2, inputting the training spectral image marked with the prosthesis type into the anti-counterfeiting recognition model to obtain a predicted prosthesis type output by the anti-counterfeiting recognition model.
Specifically, the anti-counterfeiting recognition model can output the predicted prosthesis type corresponding to the prosthesis characteristic according to the prosthesis characteristic contained in the training spectral image.
And S3, updating the parameters of the anti-counterfeiting recognition model by taking the output prosthesis type which is predicted to approach the prosthesis type marked by the training spectral image as a training target.
Specifically, the parameters of the anti-counterfeiting identification are trained and adjusted for multiple times, the anti-counterfeiting identification model with updated parameters can be obtained, and the accuracy of the output predicted prosthesis type can reach the set prediction accuracy threshold.
According to the scheme, the anti-counterfeiting recognition model can distinguish different types of prostheses after training, can cope with different types of prosthesis attacks, and improves the stability and safety of the spectrum recognition process.
The following describes the multispectral identification device provided by the embodiments of the present application, and the multispectral identification device described below and the multispectral identification method described above may be referred to correspondingly.
First, referring to fig. 3, the multispectral identification device is described, as shown in fig. 3, the multispectral identification device may include:
the image acquisition unit 100 is configured to acquire spectral images of a plurality of organisms to be detected by using a double-buffer collector, wherein two buffer collectors of the double-buffer collector are configured to sequentially acquire the spectral images, and the spectral images acquired by each buffer collector are spectral images acquired by splicing the set number of spectral images according to a set splicing manner;
the image detection unit 110 is configured to, when the target spectral images acquired by one target buffer collector of the double buffer collectors reach the set number, input the spliced spectral images to a spectral detection model to obtain a detection result of the spliced spectral images, where the spectral detection model is used to detect a spectrum included in the spectral images;
a detection result determining unit 120, configured to determine a detection result corresponding to each target spectral image according to the set stitching manner and the detection result of the stitched spectral image;
and the image identification unit 130 is configured to perform spectrum identification on each target spectral image according to a detection result corresponding to the target spectral image, so as to obtain a spectrum identification result of the target spectral image.
Optionally, the image capturing unit may include:
the storage space establishing unit is used for establishing a storage space for each buffer collector according to the set number of spectral images, and the storage space is provided with a plurality of storage positions;
and the image splicing unit is used for splicing each spectral image to the corresponding storage position in the storage space according to a set splicing mode aiming at each buffer collector to obtain the spliced spectral image.
Optionally, the multispectral identification device may further include:
the image anti-counterfeiting identification unit is used for inputting a detection result corresponding to the target spectral image into an anti-counterfeiting identification model aiming at each target spectral image before spectral identification is carried out on the target spectral image to obtain a true and false judgment result of the target spectral image;
the image recognition unit may include:
and the image identification subunit is used for performing spectrum identification on the target spectrum image with the result represented as a real organism to obtain a spectrum identification result of the target spectrum image.
Optionally, the detection result determining unit may include:
the image area determining unit is used for determining the area of each target spectral image after splicing according to the set splicing mode;
and the image area detection result determining unit is used for determining the detection result of each area of the spliced spectral image based on the detection result of the spliced spectral image, and taking the detection result of each area as the detection result of the target spectral image corresponding to the area.
Optionally, the image recognition unit may include:
the single spectrum identification unit is used for carrying out spectrum identification on the single spectrum of the target spectrum image by using a target identification threshold higher than an initial identification threshold corresponding to the single spectrum to obtain a spectrum identification result of the target spectrum image reaching the target identification threshold;
and the multispectral identification unit is used for carrying out spectrum identification on a plurality of spectrums of the target spectrum image with a plurality of spectrums of the detection result, and if the first two identified identification results are consistent, the first two identified identification results are determined as the spectrum identification results of the target spectrum image.
Optionally, the multispectral identification device may further include:
the training image acquisition unit is used for acquiring training spectral images of different prosthesis types and containing prosthesis characteristics;
the training image input unit is used for inputting the training spectral image marked with the prosthesis type into the anti-counterfeiting recognition model to obtain a predicted prosthesis type output by the anti-counterfeiting recognition model;
and the model parameter updating unit is used for updating the parameters of the anti-counterfeiting recognition model by taking the output prosthesis type which is predicted to approach the prosthesis type marked by the training spectral image as a training target.
The multispectral identification device provided by the embodiment of the application can be applied to multispectral identification equipment. Fig. 4 is a block diagram illustrating a hardware structure of the multispectral identification device, and referring to fig. 4, the hardware structure of the multispectral identification device may include: at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4;
in the embodiment of the application, the number of the processor 1, the communication interface 2, the memory 3 and the communication bus 4 is at least one, and the processor 1, the communication interface 2 and the memory 3 complete mutual communication through the communication bus 4;
the processor 1 may be a central processing unit CPU, or an application Specific Integrated circuit asic, or one or more Integrated circuits configured to implement embodiments of the present invention, etc.;
the memory 3 may include a high-speed RAM memory, and may further include a non-volatile memory (non-volatile memory) or the like, such as at least one disk memory;
wherein the memory stores a program and the processor can call the program stored in the memory, the program for:
collecting spectrum images of a plurality of organisms to be detected by using a double-buffer collector, wherein two buffer collectors of the double-buffer collector are used for collecting the spectrum images in sequence, and the spectrum image collected by each buffer collector is a spectrum image obtained by splicing the set number of spectrum images according to a set splicing mode;
when the target spectral images collected by one target buffer collector in the double buffer collectors reach the set number, inputting the spliced spectral images into a spectral detection model to obtain a detection result of the spliced spectral images, wherein the spectral detection model is used for detecting the spectrums contained in the spectral images;
determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral image;
and for each target spectral image, performing spectral recognition on the target spectral image according to a detection result corresponding to the target spectral image to obtain a spectral recognition result of the target spectral image.
Alternatively, the detailed function and the extended function of the program may be as described above.
Embodiments of the present application further provide a storage medium, where a program suitable for execution by a processor may be stored, where the program is configured to:
collecting spectrum images of a plurality of organisms to be detected by using a double-buffer collector, wherein two buffer collectors of the double-buffer collector are used for collecting the spectrum images in sequence, and the spectrum image collected by each buffer collector is a spectrum image obtained by splicing the set number of spectrum images according to a set splicing mode;
when the target spectrum images collected by one target buffer collector in the double buffer collectors reach the set number, inputting the spliced spectrum images into a spectrum detection model to obtain a detection result of the spliced spectrum images, wherein the spectrum detection model is used for detecting the spectrum contained in the spectrum images;
determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral image;
and for each target spectral image, performing spectral recognition on the target spectral image according to a detection result corresponding to the target spectral image to obtain a spectral recognition result of the target spectral image.
Alternatively, the detailed function and the extended function of the program may be as described above.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method of multispectral identification, comprising:
collecting spectrum images of a plurality of organisms to be detected by using a double-buffer collector, wherein two buffer collectors of the double-buffer collector are used for collecting the spectrum images in sequence, and the spectrum image collected by each buffer collector is a spectrum image obtained by splicing the set number of spectrum images according to a set splicing mode;
when the target spectrum images collected by one target buffer collector in the double buffer collectors reach the set number, inputting the spliced spectrum images into a spectrum detection model to obtain a detection result of the spliced spectrum images, wherein the spectrum detection model is used for detecting the spectrum contained in the spectrum images;
determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral image;
and for each target spectral image, performing spectral recognition on the target spectral image according to a detection result corresponding to the target spectral image to obtain a spectral recognition result of the target spectral image.
2. The method according to claim 1, wherein the process of stitching the spectral images comprises:
for each buffer collector:
establishing a storage space according to the set number of spectral images, wherein the storage space is provided with a plurality of storage positions;
and splicing each spectral image to a corresponding storage position in the storage space according to a set splicing mode to obtain a spliced spectral image.
3. The method of claim 1, further comprising, prior to spectrally identifying the target spectral image:
for each target spectral image, inputting a detection result corresponding to the target spectral image into an anti-counterfeiting identification model to obtain a true and false judgment result of the target spectral image;
performing spectrum identification on the target spectrum image, wherein the spectrum identification comprises the following steps:
and carrying out spectrum recognition on the target spectrum image with the result characterized as a real organism to obtain a spectrum recognition result of the target spectrum image.
4. The method according to claim 1, wherein the determining a detection result corresponding to each target spectral image according to the set stitching manner and the detection result of the stitched spectral image comprises:
determining the region of each target spectral image after splicing according to the set splicing mode;
and determining the detection result of each region of the spliced spectral image based on the detection result of the spliced spectral image, and taking the detection result of each region as the detection result of the target spectral image corresponding to the region.
5. The method according to claim 1, wherein the performing spectral recognition on each target spectral image according to the detection result corresponding to the target spectral image to obtain the spectral recognition result of the target spectral image comprises:
for a target spectrum image with a detection result only having a single spectrum, performing spectrum identification on the single spectrum of the target spectrum image by using a target identification threshold higher than an initial identification threshold corresponding to the single spectrum to obtain a spectrum identification result of the target spectrum image reaching the target identification threshold;
and performing spectrum recognition on a plurality of spectrums of the target spectrum image with a plurality of spectrums as a detection result, and determining the first two recognized recognition results as the spectrum recognition results of the target spectrum image if the first two recognized recognition results are consistent.
6. The method of claim 3, wherein the training process of the anti-counterfeit recognition model comprises:
acquiring training spectral images of different prosthesis types and containing prosthesis characteristics;
inputting the training spectral image marked with the prosthesis type into the anti-counterfeiting recognition model to obtain a predicted prosthesis type output by the anti-counterfeiting recognition model;
and updating the parameters of the anti-counterfeiting recognition model by taking the output prosthesis type which is predicted to approach the prosthesis type marked by the training spectral image as a training target.
7. A multispectral identification device, comprising:
the image acquisition unit is used for acquiring spectral images of a plurality of organisms to be detected by using a double-buffer collector, two buffer collectors of the double-buffer collector are used for sequentially acquiring the spectral images, and the spectral images acquired by each buffer collector are the spectral images acquired by splicing the set number of spectral images according to a set splicing mode;
the image detection unit is used for inputting the spliced spectral images to a spectral detection model when the target spectral images acquired by one target buffer collector in the double buffer collectors reach the set number, so as to obtain the detection result of the spliced spectral images, and the spectral detection model is used for detecting the spectrum contained in the spectral images;
the detection result determining unit is used for determining a detection result corresponding to each target spectral image according to the set splicing mode and the detection result of the spliced spectral image;
and the image identification unit is used for carrying out spectrum identification on the target spectrum image according to the detection result corresponding to the target spectrum image aiming at each target spectrum image to obtain the spectrum identification result of the target spectrum image.
8. The apparatus of claim 7, wherein the image acquisition unit comprises:
the storage space establishing unit is used for establishing a storage space for each buffer collector according to the set number of spectral images, and the storage space is provided with a plurality of storage positions;
and the image splicing unit is used for splicing each spectral image to the corresponding storage position in the storage space according to a set splicing mode aiming at each buffer collector to obtain the spliced spectral image.
9. A multispectral identification device comprising a memory and a processor;
the memory is used for storing programs;
the processor, configured to execute the program, implementing the steps of the multi-spectral identification method according to any one of claims 1 to 6.
10. A readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method for multispectral identification according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210125616.6A CN114463792B (en) | 2022-02-10 | 2022-02-10 | Multispectral identification method, multispectral identification device, multispectral identification equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210125616.6A CN114463792B (en) | 2022-02-10 | 2022-02-10 | Multispectral identification method, multispectral identification device, multispectral identification equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114463792A true CN114463792A (en) | 2022-05-10 |
| CN114463792B CN114463792B (en) | 2023-04-07 |
Family
ID=81412722
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210125616.6A Active CN114463792B (en) | 2022-02-10 | 2022-02-10 | Multispectral identification method, multispectral identification device, multispectral identification equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114463792B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040240712A1 (en) * | 2003-04-04 | 2004-12-02 | Lumidigm, Inc. | Multispectral biometric sensor |
| CN108776786A (en) * | 2018-06-04 | 2018-11-09 | 北京京东金融科技控股有限公司 | Method and apparatus for generating user's truth identification model |
| CN109214302A (en) * | 2018-08-13 | 2019-01-15 | 湖南志东科技有限公司 | One kind being based on multispectral substance identification |
| CN109583383A (en) * | 2018-11-30 | 2019-04-05 | 湖南华诺星空电子技术有限公司 | A kind of unmanned plane life detection method and system based on Multiple Source Sensor |
| CN110557527A (en) * | 2018-06-04 | 2019-12-10 | 杭州海康威视数字技术股份有限公司 | Camera and snapshot image fusion method |
| CN111126265A (en) * | 2019-12-24 | 2020-05-08 | 杭州魔点科技有限公司 | Living body detection method, living body detection device, electronic apparatus, and medium |
| CN111460419A (en) * | 2020-03-31 | 2020-07-28 | 周亚琴 | Internet of things artificial intelligence face verification method and Internet of things cloud server |
| CN113340817A (en) * | 2021-05-26 | 2021-09-03 | 奥比中光科技集团股份有限公司 | Light source spectrum and multispectral reflectivity image acquisition method and device and electronic equipment |
| CN113688806A (en) * | 2021-10-26 | 2021-11-23 | 南京智谱科技有限公司 | Infrared and visible light image fused multispectral target detection method and system |
-
2022
- 2022-02-10 CN CN202210125616.6A patent/CN114463792B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040240712A1 (en) * | 2003-04-04 | 2004-12-02 | Lumidigm, Inc. | Multispectral biometric sensor |
| CN108776786A (en) * | 2018-06-04 | 2018-11-09 | 北京京东金融科技控股有限公司 | Method and apparatus for generating user's truth identification model |
| CN110557527A (en) * | 2018-06-04 | 2019-12-10 | 杭州海康威视数字技术股份有限公司 | Camera and snapshot image fusion method |
| CN109214302A (en) * | 2018-08-13 | 2019-01-15 | 湖南志东科技有限公司 | One kind being based on multispectral substance identification |
| CN109583383A (en) * | 2018-11-30 | 2019-04-05 | 湖南华诺星空电子技术有限公司 | A kind of unmanned plane life detection method and system based on Multiple Source Sensor |
| CN111126265A (en) * | 2019-12-24 | 2020-05-08 | 杭州魔点科技有限公司 | Living body detection method, living body detection device, electronic apparatus, and medium |
| CN111460419A (en) * | 2020-03-31 | 2020-07-28 | 周亚琴 | Internet of things artificial intelligence face verification method and Internet of things cloud server |
| CN113340817A (en) * | 2021-05-26 | 2021-09-03 | 奥比中光科技集团股份有限公司 | Light source spectrum and multispectral reflectivity image acquisition method and device and electronic equipment |
| CN113688806A (en) * | 2021-10-26 | 2021-11-23 | 南京智谱科技有限公司 | Infrared and visible light image fused multispectral target detection method and system |
Non-Patent Citations (3)
| Title |
|---|
| MD OSMAN GANI ET AL.: "Multispectral object detection with deep learning", 《INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE IN COMMUNICATIONS AND BUSINESS ANALYSIS》 * |
| 李元耀: "基于LabVIEW的自动光谱检测系统及其应用", 《科技视界》 * |
| 蒋英春: "基于成像传感器特征的相机源识别", 《计算机工程与应用》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114463792B (en) | 2023-04-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110751022B (en) | Urban pet activity track monitoring method based on image recognition and related equipment | |
| CN109145742B (en) | Pedestrian identification method and system | |
| US10558841B2 (en) | Method and apparatus for recognizing fingerprint ridge point | |
| CN105868613A (en) | Biometric feature recognition method, biometric feature recognition device and mobile terminal | |
| CN102945366A (en) | Method and device for face recognition | |
| CN111462379A (en) | Access control management method, system and medium containing palm vein and face recognition | |
| CN104598888B (en) | A kind of recognition methods of face gender | |
| CN107918773B (en) | Face living body detection method and device and electronic equipment | |
| CN101681497A (en) | Vein pattern management system, vein pattern registration device, vein pattern authentication device, vein pattern registration method, vein pattern authentication method, program, and vein data struc | |
| CN107786867A (en) | Image identification method and system based on deep learning architecture | |
| CN108171138A (en) | A kind of biological information acquisition methods and device | |
| CN106991448A (en) | A kind of portrait compares processing method | |
| CN110322470A (en) | Action recognition device, action recognition method and recording medium | |
| CN114387548A (en) | Video and liveness detection method, system, device, storage medium and program product | |
| CN111488798B (en) | Fingerprint identification method, fingerprint identification device, electronic equipment and storage medium | |
| CN111079687A (en) | Certificate camouflage identification method, device, equipment and storage medium | |
| CN108875497B (en) | Living body detection method, living body detection device and computer storage medium | |
| CN111027534A (en) | Compact double-license-plate detection method and device | |
| CN108875549A (en) | Image-recognizing method, device, system and computer storage medium | |
| TWI770927B (en) | Fingerprint identification method, storage medium, and terminal | |
| CN114463792B (en) | Multispectral identification method, multispectral identification device, multispectral identification equipment and readable storage medium | |
| KR101336834B1 (en) | Usb iris recognition device | |
| CN112580531A (en) | Method and system for identifying and detecting true and false license plates | |
| CN114463776A (en) | Fall identification method, device, equipment and storage medium | |
| CN110390353B (en) | Biological identification method and system based on image processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |