WO2022222904A1

WO2022222904A1 - Image verification method and system, and storage medium

Info

Publication number: WO2022222904A1
Application number: PCT/CN2022/087565
Authority: WO
Inventors: 张明文; 张天明; 赵宁宁
Original assignee: 北京嘀嘀无限科技发展有限公司
Priority date: 2021-04-20
Filing date: 2022-04-19
Publication date: 2022-10-27

Abstract

Disclosed in the embodiments of the present description are an image verification method, system and apparatus, and a storage medium. The image verification method comprises: instructing a terminal device to emit at least two beams of colored light to a target; during the process of emitting the at least two beams of colored light, instructing the terminal device to collect a plurality of images of the target; selecting, from the plurality of images, at least two target images that respectively correspond to the at least two beams of colored light; on the basis of the at least two target images, determining a target verification code; and performing image authenticity verification by comparing the target verification code with a reference verification code corresponding to the at least two beams of colored light.

Description

Image verification method, system and storage medium

cross reference

This application claims priority to Chinese Patent Application No. 202110423967.0 filed on April 20, 2021 and Chinese Patent Application No. 202210246335.6 filed on March 14, 2022, the entire contents of which are incorporated herein by reference.

technical field

The present specification relates to the technical field of image processing, and in particular, to a method, system and storage medium for image verification.

Background technique

Image verification is a technology for identification or verification based on images collected by image acquisition devices, and is widely used in application scenarios such as identity verification, authority verification, and target recognition. In practical applications, there may be a problem of illegal verification through camera hijacking and other methods. In addition, due to environmental conditions (for example, low ambient brightness, the presence of colored light in the environment, etc.) may have an impact on the image acquisition process, and the impact of light on different areas of the target may be different, resulting in large errors in the verification results. . In order to ensure the security of target recognition, it is necessary to determine the authenticity of the target image.

Therefore, it is necessary to provide an image verification method, system and storage medium to determine the authenticity of the target image and improve the accuracy and security of image verification.

SUMMARY OF THE INVENTION

One of the embodiments of this specification provides an image verification method, the method includes: instructing a terminal device to emit at least two colored light beams to a target; during the process of transmitting the at least two colored light beams, instructing the terminal device to collect the multiple images of the target; selecting at least two target images respectively corresponding to the at least two colored rays from the multiple images; determining a target verification code based on the at least two target images; verifying the target by comparing the The code and the reference verification code corresponding to the at least two colored light beams are used to verify the authenticity of the image.

One of the embodiments of the present specification provides an image verification system, the system includes: a light emission module for instructing a terminal device to emit at least two colored light beams to a target. The image verification system further includes an image acquisition module for instructing the terminal device to collect multiple images of the target during the process of emitting the at least two colored light beams and to select the at least two images from the multiple images. At least two target images corresponding to the bundles of colored rays respectively. The image verification system further includes a target verification code determination module for determining a target verification code based on the at least two target images. The image verification system further includes an image authenticity determination module for performing image authenticity verification by comparing the target verification code with the reference verification codes corresponding to the at least two colored light beams.

One of the embodiments of this specification provides a computer-readable storage medium, where the storage medium stores computer instructions, and when the computer reads the computer instructions in the storage medium, the computer executes the image verification method disclosed in this specification.

In this manual, the problem of camera hijacking can be effectively solved by instructing the terminal to emit colored light and collecting corresponding images for image verification. In addition, the target image is divided into a reference image and a verification image, and the target verification code is determined based on the difference between them (that is, the verification code is determined by self-comparison or self-query); correspondingly, the multi-colored rays are divided into reference rays. And the verification light, determine the reference verification code based on the difference between each other, and then verify by comparing whether the target verification code and the reference verification code are consistent, which can eliminate the influence of environmental conditions and improve the accuracy of the verification result. Further, in addition to the color (hue) parameters of the light or image, the brightness and/or saturation are also considered. Accordingly, the number of digits in the verification code is more, which greatly reduces the probability of being broken, and the verification accuracy and Security is also higher. Furthermore, dividing the reference image and the verification image into multiple image blocks, and performing subsequent similarity processing and analysis based on the multiple image blocks through the transformer structure can effectively solve the problem that light affects different areas of the target differently. Problems, improve the model's refined processing ability, thereby improving the verification effect.

Description of drawings

The present specification will be further described by way of example embodiments, which will be described in detail with reference to the accompanying drawings. These examples are not limiting, and in these examples, the same numbers refer to the same structures, wherein:

1 is a schematic diagram of an application scenario of an exemplary image verification system according to some embodiments of the present specification;

2 is a block diagram of an exemplary image verification system according to some embodiments of the present specification;

3 is a flowchart of an exemplary image verification process according to some embodiments of the present specification;

FIG. 4 is an exemplary flowchart of determining a target verification code based on a target image according to some embodiments of the present specification;

FIG. 5 is a schematic diagram illustrating an exemplary determination of the difference between a reference image and a verification image according to some embodiments of the present specification;

6 is a schematic diagram of an exemplary image verification process according to some embodiments of the present specification;

7 is a schematic diagram of an exemplary alignment model shown in accordance with some embodiments of the present specification;

8A and 8B are schematic diagrams of determining the similarity between a verification image and a reference image according to some embodiments of the present specification;

9 is a schematic diagram of constructing spatial attention according to some embodiments of the present specification;

10 is a flowchart of another exemplary image verification process according to some embodiments of the present specification;

Figure 11 is a schematic diagram of an exemplary lighting sequence according to some embodiments of the present specification;

FIG. 12 is a schematic structural diagram of an exemplary color verification model according to some embodiments of the present specification.

Detailed ways

In order to illustrate the technical solutions of the embodiments of the present specification more clearly, the following briefly introduces the accompanying drawings used in the description of the embodiments. Obviously, the accompanying drawings in the following description are only some examples or embodiments of the present specification. For those of ordinary skill in the art, without any creative effort, the present specification can also be applied to the present specification according to these drawings. other similar situations. Unless obvious from the locale or otherwise specified, the same reference numbers in the figures represent the same structure or operation.

It is to be understood that "system", "device", "unit" and/or "module" as used herein is a method used to distinguish different components, elements, parts, parts or assemblies at different levels. However, other words may be replaced by other expressions if they serve the same purpose.

As shown in the specification and claims, unless the context clearly dictates otherwise, the words "a", "an", "an" and/or "the" are not intended to be specific in the singular and may include the plural. Generally speaking, the terms "comprising" and "comprising" only imply that the clearly identified steps and elements are included, and these steps and elements do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

Flowcharts are used in this specification to illustrate operations performed by a system according to an embodiment of this specification. It should be understood that the preceding or following operations are not necessarily performed in the exact order. Instead, the various steps can be processed in reverse order or simultaneously. At the same time, other actions can be added to these procedures, or a step or steps can be removed from these procedures.

In this specification, "images" may include still images (eg, individual images, video frames, etc.) or dynamic images (eg, video).

Target recognition (also referred to as "target verification") is a technology for biometric identification based on target objects acquired by image acquisition devices. In some embodiments, the target object (also referred to as a "target") may be a human face, a fingerprint, a palm print, a pupil, or the like. In some embodiments, object recognition may be applied to authorization verification. For example, access control authority authentication, account payment authority authentication, etc. In some embodiments, target recognition may also be used for authentication. For example, employee attendance authentication, personal registration identity security authentication, etc. Merely as an example, the target identification may be based on matching of the target image captured by the image capture device in real time and the pre-acquired biometric features, thereby verifying the target identity.

However, image capture devices can be hacked or hijacked, and attackers can upload fake target images for authentication. For example, attacker A can directly upload the face image of user B after attacking or hijacking the image acquisition device. The target recognition system (also referred to as an "image verification system") performs face recognition based on user B's face image and pre-acquired user B's face biometrics, thereby passing user B's identity verification.

Therefore, in order to ensure the security of target recognition, it is necessary to determine the authenticity of the target image, that is, to determine that the target image is collected in real time by the image acquisition device during the target recognition process.

FIG. 1 is a schematic diagram of an application scenario of an exemplary image verification system according to some embodiments of the present specification. In some embodiments, the image verification system 100 may be applied to scenarios such as facial recognition verification, object verification (eg, vehicle verification, document verification, etc.). In some embodiments, as shown in FIG. 1, an image verification system 100 (also referred to as an "object recognition system") may include a server 110, a network 120, a terminal device (also referred to as a "terminal") 130 and storage device 140 .

Server 110 may be a single server or a group of servers. The server group may be centralized or distributed (eg, server 110 may be a distributed system). In some embodiments, server 110 may be local or remote. For example, the server 110 may access information and/or data stored in the terminal device 130 and/or the storage device 140 through the network 120 . For another example, the server 110 may be directly connected to the terminal device 130 and/or the storage device 140 to access stored information and/or data. In some embodiments, server 110 may be implemented on a cloud platform. By way of example only, cloud platforms may include private clouds, public clouds, hybrid clouds, community clouds, distributed clouds, internal clouds, multi-tier clouds, etc., or any combination thereof.

In some embodiments, server 110 may include processing device 112 for implementing the example methods and/or systems described in this specification. For example, the processing device 112 may instruct the terminal device 130 to emit at least two colored rays to the target, and instruct the terminal device 130 to collect multiple images of the target during the process of emitting the colored rays. The processing device 112 may select at least two (may also be referred to as "multiple") target images corresponding to the at least two colored rays from the plurality of images, and determine the target verification code based on the at least two target images. Further, the processing device 112 may perform image authenticity verification by comparing the target verification code with the reference verification codes corresponding to the at least two colored rays. For another example, the processing device 112 may determine the first color relationship and the second color relationship, and determine the authenticity of the plurality of target images, and the like. During processing, processing device 112 may obtain data (eg, instructions) from other components of object recognition system 100 (eg, storage device 140 , terminal 130 ) directly or through network 120 and/or send the processed data to others components for storage or display.

In some embodiments, server 110 and/or processing device 112 may be implemented by a computing device, eg, a computing device including a processor, memory, network interface, communication interface, display device, and the like.

Network 120 may facilitate the exchange of data and/or information. In some embodiments, components in image verification system 100 (eg, server 110 , terminal device 130 , storage device 140 ) may send information and/or data to other components in image verification system 100 via network 120 . For example, the server 110 may acquire data from the terminal device 130 and/or the storage device 140 through the network 120 . In some embodiments, the network 120 may be any one or more of a wired network or a wireless network. For example, the network 120 may include a cable network, a fiber optic network, a telecommunications network, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN) , Bluetooth network, ZigBee network (ZigBee), Near Field Communication (NFC) network, etc. or any combination thereof. In some embodiments, the network connection between the various parts in the object recognition system 100 may adopt one of the above-mentioned manners, or may adopt multiple manners. In some embodiments, the network 120 may be of various topologies such as point-to-point, shared, centralized, or a combination of topologies. In some embodiments, network 120 may include one or more network access points. For example, network 120 may include wired or wireless network access points, eg, base stations and/or network switching points 120-1, 120-2, . . . , through which one or more components of image verification system 100 are used. A network 120 may be connected to exchange data and/or information.

The terminal device 130 may capture an image of the target. In some embodiments, the terminal device 130 may include an image capture device 130-1, a mobile device 130-2, a tablet computer 130-3, a laptop computer 130-4, etc., or any combination thereof. In some embodiments, the image capture device 130-1 may include a camera, a camera, an image sensor, etc., or any combination thereof. In some embodiments, the image capturing device 130-1 may photograph the target object and acquire multiple target images. In some embodiments, the terminal device 130 may include any device having an image capture function. In some embodiments, the terminal device 130 may emit light (eg, colored light) to the target during image acquisition. In some embodiments, the terminal device 130 may emit light through its own light emitting elements (eg, screen, LED lights, etc.). In some embodiments, the terminal device 130 may emit light through a light-emitting device (eg, an external LED lamp, a light-emitting diode, etc.) connected to it. In some embodiments, the terminal device 130 may communicate with the processing device 112 through the network 120 and transmit the image of the captured object to the processing device 112 .

Storage device 140 may be used to store data (eg, lighting sequences, multiple target images, first color relationships, second color relationships, etc.) and/or instructions. In some embodiments, the storage device 140 may store data obtained from the server 110 and/or the terminal device 130 (eg, images captured by the terminal device 130 ). In some embodiments, storage device 150 may store data and/or instructions for execution or use by server 110 (eg, data and/or instructions for implementing the example methods described herein). The storage device 140 may include one or more storage components, and each storage component may be an independent device or a part of other devices. In some embodiments, storage device 140 may include random access memory (RAM), read only memory (ROM), mass storage, removable memory, volatile read-write memory, the like, or any combination thereof. In some embodiments, storage device 140 may be implemented on a cloud platform. By way of example only, cloud platforms may include private clouds, public clouds, hybrid clouds, community clouds, distributed clouds, internal clouds, multi-tier clouds, etc., or any combination thereof. In some embodiments, storage device 140 may be integrated or included in one or more other components of object recognition system 100 (eg, processing device 112, terminal 130, or other possible components).

In some embodiments, the storage device 140 may be connected to the network 120 to enable communication with other components in the image verification system 100 (eg, the server 110, the terminal device 130). In some embodiments, the storage device 140 may directly connect or communicate with other components of the image verification system 100 (eg, the server 110, the terminal device 130). In some embodiments, storage device 140 may be part of server 110 or terminal device 130 .

It should be noted that the above description of the target recognition system is only for the convenience of description, and does not limit the description to the scope of the illustrated embodiments. It will be appreciated that those skilled in the art, with an understanding of the principles of the system, may make changes to the system and its components without departing from such principles.

FIG. 2 is a block diagram of an exemplary image verification system shown in accordance with some embodiments of the present specification. In some embodiments, image verification system 200 may be implemented by processing device 112 . In some embodiments, the image verification system 200 may include a light emission module 210, an image acquisition module (which may also be referred to as a "target image acquisition module") 220, a target verification code determination module 230, and an image authenticity determination module (which may also be referred to as Referred to as an "authentication module") 240.

The light emission module 210 may instruct the terminal device 130 to emit at least two colored lights (at least two colored lights may also be referred to as "lighting sequences") to the target. In some embodiments, the light emission module 210 may instruct the terminal device 130 to randomly emit at least two colored light beams. In some embodiments, the light emission module 210 may instruct the terminal device 130 to emit at least two colored lights based on a preset rule. For more details about the emitted light, please refer to step 310 and its related description, which will not be repeated here.

The image acquisition module 220 may instruct the terminal device 130 to acquire multiple images of the target during the process of emitting at least two colored light beams. In some embodiments, the image acquisition module 220 may select at least two target images respectively corresponding to at least two colored rays from the plurality of images. For more content about image acquisition, refer to

steps

320 and 330 and their related descriptions, which will not be repeated here.

In some embodiments, the target image acquisition module may be configured to acquire multiple target images, the shooting time of the multiple target images has a corresponding relationship with the irradiation time of the multiple lightings in the lighting sequence irradiating the target object, and the multiple lightings have a corresponding relationship. colors, the plurality of colors include at least one reference color and at least one verification color, and the plurality of target images include at least one verification image (also referred to as "verification image") and at least one reference image, and at least one reference image Each of the images corresponds to one of the at least one reference color, and each of the at least one verification image corresponds to one of the at least one verification color. In some embodiments, one or more of the at least one reference color is the same as one or more of the at least one verification color. For more information on reference color and verification color, please refer to Figure 11 and its related descriptions.

The target verification code determination module 230 may determine the target verification code based on at least two target images. In some embodiments, the target verification code determination module 230 may determine the reference image and the verification image based on the at least two target images, and determine the target verification code based on differences (eg, differences in image parameters) between the reference image and the verification image. In some embodiments, the target verification code determination module 230 may determine one of the at least two target images as a verification image, and use the other image of the at least two target images as a reference image. For more content about the target verification code, refer to step 340 and its related description, which will not be repeated here.

The image authenticity determination module 240 may perform image authenticity verification by comparing the target verification code with the reference verification codes corresponding to at least two colored rays. If the target verification code is consistent with the reference verification code, it means that the image authenticity verification has passed, and further indicates that the target's identity verification, authority verification and/or authenticity verification have passed. If the target verification code is inconsistent with the reference verification code, the image authenticity verification fails. For more details about the determination of the authenticity of the image, reference may be made to step 350 and its related description, which will not be repeated here.

In some embodiments, the target recognition system may further include a first color relationship determination module, a second color relationship determination module, and a model acquisition module.

The first color relationship determination module may be configured to determine, for each of the at least one reference image, the first color relationship between the reference image and each verification image.

In some embodiments, the first color relationship determination module may extract the reference color feature of the reference image and the verification color feature of each verification image; and determine the reference image and each verification image based on the reference color feature and the verification color feature The first color relationship of .

In some embodiments, each of the at least one reference image and each of the at least one verification image form at least one pair of images, and for each pair of the at least one pair of images, the first color relationship is determined The module can process the image pair based on the color verification model, and determine the first color relationship between the reference image and the verification image in the image pair, and the color verification model is a machine learning model with preset parameters. In some embodiments, the color verification model includes a color feature extraction layer and a color relationship determination layer, where the color feature extraction layer is used to extract color features of the image pair; the color relationship determination layer determines the reference image in the image pair based on the color features of the image pair and verify the first color relationship of the image.

The second color relationship determination module may be configured to determine, for each of the at least one reference color, a second color relationship between the reference color and each verification color.

In some embodiments, the verification module may be configured to determine a target verification code based on the at least one first color relationship and a reference verification code based on the at least one second color relationship, and determine the target verification code based on the target verification code and the reference verification code. authenticity. For more details about determining the authenticity of multiple target images based on the target verification code and the reference verification code, please refer to FIG. 10 and related descriptions.

The model acquisition module can be used to acquire color verification models. The preset parameters of the color verification model are obtained through end-to-end training. In some embodiments, the training process includes acquiring a plurality of training samples, each of the plurality of training samples including a sample image pair and a sample label, the sample label indicating whether the sample images in the sample image pair are illuminated by light of the same color and the initial color verification model is trained based on multiple training samples, and the preset parameters of the color verification model are determined.

For more detailed description of the target image acquisition module, the first color relationship determination module, the second color relationship determination module, the verification module and the model acquisition module, please refer to FIGS.

It should be understood that the image verification system 200 and its modules shown in FIG. 2 can be implemented in various ways, for example, implemented by hardware, software, or a combination of software and hardware. The system and its modules of this specification can be implemented not only by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. , can also be implemented by, for example, software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software (eg, firmware).

It should be noted that the above description of the image verification system 200 and its modules is only for the convenience of description, and does not limit the description to the scope of the illustrated embodiments. It can be understood that for those skilled in the art, after understanding the principle of the system, various modules may be combined arbitrarily, or a subsystem may be formed to connect with other modules without departing from the principle. In some embodiments, the light emission module 210 , the image acquisition module 220 , the target verification code determination module 230 and the image authenticity determination module 240 disclosed in FIG. 2 may be different modules in a system, or may be one module to implement the above function of two or more modules. For example, each module may share one storage module, and each module may also have its own storage module. Such deformations are all within the protection scope of this specification.

3 is a flowchart of an exemplary image verification process shown in accordance with some embodiments of the present specification. In some embodiments, process 300 may be performed by

image verification system

100 or 200 . For example, the process 300 can be stored in a storage device (eg, the storage device 140, a storage unit of the processing device 112) in the form of a program or an instruction, and when the processing device 112 or the module shown in FIG. 2 executes the program or the instruction, it can be implemented Process 300. In some embodiments, process 300 may be accomplished with one or more additional operations not described below, and/or without one or more operations discussed below. Additionally, the order of operations shown in FIG. 3 is not limiting.

Step 310, instructing the terminal device 130 to emit at least two colored light beams to the target. In some embodiments, step 310 may be performed by the light emission module 210 .

Target refers to the object that requires image validation. In some embodiments, the target may be the face of a user requiring authentication and/or authorization. For example, in an online taxi-hailing scenario, the taxi-hailing platform needs to verify whether the driver who takes the order is a registered driver user approved by the platform, and the target is the face of the driver who takes the order. For another example, in a payment scenario, the payment system needs to verify the payment authority of the payer, and the target is the payer's face. In some embodiments, the target may be an object requiring authenticity verification. For example, in a vehicle verification scenario, the vehicle verification platform needs to verify whether the vehicle is a real vehicle, and the target can be the vehicle to be verified.

In some embodiments, the terminal device 130 may emit colored light through its own light emitting elements (eg, screen, LED lights, etc.). In some embodiments, the terminal device 130 may emit colored light through a light-emitting device (eg, an external LED lamp, a light-emitting diode, etc.) connected thereto.

In some embodiments, the terminal device 130 may sequentially emit at least two colored light beams in chronological order. In some embodiments, the durations of each colored light beam may be the same or different.

In some embodiments, in order to ensure that the image of the target can be captured within the duration of the colored light, and at the same time to avoid affecting the user experience due to the long duration, the duration of each colored light needs to be set within a preset range. In some embodiments, the preset range may be 100-500 milliseconds. In some embodiments, the preset range may be 150-450 milliseconds. In some embodiments, the preset range may be 200-400 milliseconds. In some embodiments, the preset range may be 250-350 milliseconds. In some embodiments, the preset range may be 300 milliseconds. In some embodiments, the preset range may be 250 milliseconds. In some embodiments, the preset range may be 200 milliseconds.

In some embodiments, different conditions may correspond to different durations. For example, if the ambient brightness is low, the duration may be relatively long; conversely, if the ambient brightness is high, the duration may be relatively short.

In some embodiments, the terminal device 130 may emit multiple colored light beams, wherein at least two colored light beams have different colors. By setting colored lights of different colors, the subsequently collected images can have a certain degree of distinction to ensure the effect of authenticity verification, and at the same time, to ensure that the imaging effect of the target (for example, the face) is better. In some embodiments, the light parameters of the colored light can be represented by the HSV color model, including hue (H), saturation (S), and brightness (V). Correspondingly, at least two of the multiple colored light beams have different hues (which may reflect the difference in color).

In some embodiments, the respective hues of the at least two colored lights may be set randomly, and at least one of brightness or saturation may be a variable value. For example, the hue corresponding to the at least two colored rays may be set to a random value between 0 and 1, and one of brightness or saturation may be set to a variable value between 0 and 1. For another example, the respective hues of the at least two colored lights may be set to random values between 0 and 1, and the brightness and saturation may both be set to varying values between 0 and 1. Through the random setting of the hue, it can be basically ensured that the colors of at least two of the multiple colored lights are different. In addition, through the change of brightness or saturation, the distinction of multiple colored lights can be made more obvious, and the number of digits of the verification code involved in the subsequent verification process can also be increased, thereby ensuring the accuracy and security of image verification. It should be noted that, in this specification, "variation value" is relative to "fixed value", for example, a change value between 0 and 1 can be understood as a random change value between 0 and 1.

In some embodiments, the at least two colored light beams respectively correspond to different hues, and at least one of brightness or saturation is different. For example, the respective hues of the at least two colored rays may be set to a value between 0 and 1 different from each other, and one of brightness or saturation may be set to a value between 0 and 1 different from each other. For another example, the respective hues of the at least two colored rays may be set to be values between 0 and 1 that are different from each other, and both the brightness and the saturation may be set to be values between 0 and 1 that are different from each other. The hue of at least two colored lights is different, which can ensure that the colors of the colored lights are different, and at least one of the brightness or saturation is different. On the basis of different colors, the distinction between the colored lights can be further strengthened, and the subsequent verification process can be increased. The number of digits involved in the verification code will correspondingly improve the accuracy and security of subsequent image verification.

In some embodiments, the hue, saturation and brightness corresponding to the at least two colored light beams may be randomly set. For example, the hue, saturation, and brightness corresponding to at least two colored rays may be set to random values between 0 and 1. In some embodiments, the hue and brightness corresponding to the at least two colored lights can be set randomly, and the saturation is a fixed value. For example, the hue and brightness corresponding to the at least two colored rays may be set to random values between 0 and 1, and the saturation may be set to a fixed value (eg, 1) between 0 and 1. In some embodiments, the hue and saturation corresponding to the at least two colored lights can be randomly set, and the brightness is a fixed value. For example, the hue and saturation corresponding to the at least two colored rays may be set to random values between 0 and 1, and the brightness may be set to a fixed value (eg, 1) between 0 and 1.

In some embodiments, the light parameters of the colored light can also be represented by other color models, for example, RGB, Lαβ, CMYK, NTSC, LMS, YcbCr, etc., which are not limited in this specification.

In some embodiments, the light emission module 210 may instruct the terminal device 130 to randomly emit at least two colored light beams. In some embodiments, the light emission module 210 may instruct the terminal device 130 to emit at least two colored lights based on a preset rule. For example, different verification scenarios can correspond to different colored lights. For another example, different environmental conditions (eg, ambient brightness) may correspond to different colored lights. For another example, different objects (eg, faces, vehicles, etc.) may correspond to different colored lights.

In some embodiments, the relevant parameters of the colored light can be stored in the storage device 140, and the light emission module 210 can obtain the relevant parameters of the colored light from the storage device 140, and instruct the terminal device 130 to emit the corresponding colored light.

Step 320, instructing the terminal device 130 to collect multiple images of the target during the process of emitting at least two colored light beams. In some embodiments, step 320 may be performed by image acquisition module 220 .

In some embodiments, during the emission process of each colored light beam (ie, within the duration of each colored light beam), the terminal device 130 may perform one or more image acquisitions on the target. For example, as shown in FIG. 6 , the light emission module 210 instructs the terminal device 130 to emit light 1, light 2, light 3, light 4, light 5 and light 6, a total of 6 colored light beams. During the emission of each colored light beam, the image acquisition module 220 instructs the terminal device 130 to perform one or more image acquisitions on the target to obtain one or more images of the target (eg, the initial image shown in FIG. 6 ).

In some embodiments, during the emission process of each colored light beam, the number of times of image acquisition may be the default value of the system, or may be adjusted according to different situations. For example, if the ambient brightness is low, the number of acquisitions may be relatively large; conversely, if the ambient brightness is high, the number of acquisitions may be relatively small.

In some embodiments, the duration of each image acquisition may be a system default value, or may be adjusted according to different situations. For example, if the ambient brightness is low, the duration may be correspondingly longer; conversely, if the ambient brightness is high, the duration may be relatively short.

In some embodiments, the time interval between multiple image acquisitions may be a system default value (eg, 2 milliseconds, 5 milliseconds, 7 milliseconds, 8 milliseconds, 10 milliseconds, 20 milliseconds, etc.), or may be adjusted according to different situations.

By collecting multiple initial images during the emission of each colored light, even if the quality of one or more images is poor or the target is not captured, it can be ensured that the available target images can be screened and the subsequent images can be obtained. The verification process is proceeding normally.

In some embodiments, the format of the image may include Joint Photographic Experts Group (JPEG), Tagged Image File Format (TIFF), Graphics Interchange Format (GIF), Kodak Flash PiX (FPX), Digital Imaging and Communications in Medicine (DICOM) etc. or any combination thereof. In some embodiments, the image may be a two-dimensional image, a three-dimensional image, a four-dimensional image, or the like.

Step 330: Select at least two target images corresponding to at least two colored rays from the multiple images. In some embodiments, step 330 may be performed by image acquisition module 220 .

In some embodiments, for each bundle of colored rays, the image acquisition module 220 may select an image containing the target from the multiple initial images corresponding to the bundle of colored rays as the target image. For example, as shown in FIG. 6 , for 6 colored rays, the image acquisition module 220 can respectively determine 6 target images S1 , S2 , S3 , S4 , S5 and S6 .

In some embodiments, for each colored light beam, the image acquisition module 220 may randomly select an image containing a target from a plurality of initial images corresponding to the colored light beam as the target image. In some embodiments, the image acquisition module 220 may select an image that satisfies the preset requirements from the multiple initial images corresponding to the bundle of colored rays as the target image. In some embodiments, the preset requirements may include that the image quality meets the quality requirements (eg, brightness greater than a preset brightness threshold, sharpness greater than a preset sharpness threshold, contrast greater than a preset contrast threshold, etc.), the position of the target in the image The requirements are met (for example, it is located at or basically in the middle of the image field of view), and the size of the target in the image meets the requirements (for example, the size of the target reaches 79%, 80%, 90% or 95% of the overall size of the image, etc.), etc. or any combination thereof.

In some embodiments, after determining at least two target images respectively corresponding to the at least two colored rays, the image acquisition module 220 may preprocess the target images to improve image quality, thereby improving the accuracy of subsequent image verification.

In some embodiments, the preprocessing may include texture uniform processing. It can be understood that during the image acquisition process, the distance, angle, and background of the terminal device 130 and the target may vary, and accordingly, the texture of each target image may be different. The texture uniform processing can make the texture of each target image the same or basically the same, reduce the interference of texture features, and improve the efficiency and accuracy of subsequent image verification.

In some embodiments, the preprocessing may further include operations such as image denoising, image enhancement, etc., which can improve the effect of subsequent image verification, which is not limited in this specification.

Step 340: Determine a target verification code based on the at least two target images. In some embodiments, step 340 may be performed by target verification code determination module 230 .

In some embodiments, the target verification code determination module 230 may determine the reference image and the verification image based on the at least two target images, and determine the target verification code based on differences (eg, differences in image parameters) between the reference image and the verification image. For more details about determining the target verification code based on at least two target images, reference may be made to FIG. 4 and FIG. 5 and related descriptions, and details are not repeated here.

In some embodiments, the target verification code determination module 230 may also determine the target verification code based on the first color relationship between the reference image and each verification image. For more content about determining the target verification code based on the first color relationship, please refer to FIG. 10 and related descriptions.

In step 350, the authenticity of the image is verified by comparing the target verification code with the reference verification codes corresponding to at least two beams of colored light. In some embodiments, step 350 may be performed by image authenticity determination module 240 .

In some embodiments, the reference verification code may be determined based on light parameters of at least two colored light beams. In some embodiments, the reference verification code is determined in a manner similar to the determination of the target verification code. For example, the reference light and the verification light are determined based on at least two colored light rays, and the reference verification code is determined based on the difference between the reference light and the verification light (eg, the difference in light parameters). For more description, reference may be made to FIG. 4 and FIG. 5 , which will not be repeated here.

In some embodiments, the reference verification code may also be determined based on the reference color and the second color relationship of each verification color. For more content about determining the reference verification code based on the second color relationship, please refer to FIG. 10 and related descriptions.

In combination with step 310, in some embodiments, the reference verification code may be determined offline and stored in the storage device 140 together with at least two colored rays (or light parameters) corresponding to the reference verification code. After the processing device 112 instructs the terminal device 130 to emit at least two colored lights and collect corresponding images, the processing device 112 can read the reference verification codes corresponding to the at least two colored lights from the storage device 140 and perform image authenticity verification. In some embodiments, processing device 112 may determine the reference verification code at the same time as or after instructing terminal device 130 to emit at least two colored light beams.

In some embodiments, if the target verification code is consistent with the reference verification code, it indicates that the authenticity verification of the image has passed, and further indicates that the identity verification, authority verification and/or authenticity verification of the target has passed. If the target verification code is inconsistent with the reference verification code, the image authenticity verification fails.

In some embodiments, the processing device 112 may instruct the terminal device 130 to issue a pass or fail related alert or notification. In some embodiments, reminders or notifications may be presented in text, images, audio, video, and the like.

It should be noted that the above description about the process 300 is only for example and illustration, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to the process 300 under the guidance of this specification. However, these corrections and changes are still within the scope of this specification.

FIG. 4 is an exemplary flowchart of determining a target verification code based on a target image according to some embodiments of the present specification. In some embodiments, process 400 may be performed by

image verification system

100 or 200 . For example, the process 400 can be stored in a storage device (eg, the storage device 140, a storage unit of the processing device 112) in the form of a program or an instruction, and when the processing device 112 or the module shown in FIG. 2 executes the program or the instruction, it can be implemented Process 400. In some embodiments, process 400 may be accomplished with one or more additional operations not described below, and/or without one or more operations discussed below. Additionally, the order of operations shown in FIG. 4 is not limiting.

Step 410: Determine one of the at least two target images as a verification image, and use the other image in the at least two target images as a reference image. In some embodiments, step 410 may be accomplished by the target verification code determination module 230 .

In some embodiments, the target image corresponding to the verification ray may be determined as the verification image, and the target image corresponding to the reference ray may be determined as the reference image. In some embodiments, the verification image/verification ray and the reference image/reference ray may be randomly set. For example, as shown in FIG. 6 , light 6 can be used as the verification light, and the target image S6 corresponding to light 6 can be used as the verification image; light 1, light 2, light 3, light 4, and light 5 can be used as The target images S1 , S2 , S3 , S4 , and S5 corresponding to the ray 1 , the ray 2 , the ray 3 , the ray 4 , and the ray 5 respectively serve as the reference images.

Step 420, for each of the reference images, determine the difference between the reference image and the verification image. In some embodiments, step 420 may be accomplished by the target verification code determination module 230 .

In some embodiments, the difference may reflect the difference between the reference image and the verification image in various dimensions or in various aspects (eg, spatial location, gray value, gradient value, resolution, etc.). In some embodiments, in conjunction with step 310, the light parameters of the colored light can be represented by the HSV color model. Correspondingly, the image parameters of the target image corresponding to the colored light can also be represented by the HSV color model. In this case, the difference can reflect the difference in hue, brightness, and saturation between the reference image and the verification image. For example, for "Hue", if the color tone of the reference image and the verification image are the same, the "Hue" difference is 1, otherwise it is 0; for "Brightness", if the brightness of the reference image and the verification image are the same, then The difference is 1, otherwise it is 0; for "Saturation", if the saturation of the reference image and the verification image are the same, the "Saturation" difference is 1, otherwise it is 0. In some embodiments, the combined difference of the "hue" difference, the "brightness" difference and the "saturation" difference may be used as the difference between the reference image and the verification image. For example, if the reference image and the verification image are different in hue, but the brightness and saturation are the same, the difference may be 011. In some embodiments, the differences may be represented in numerical values, vectors, matrices, or the like.

In some embodiments, the reference image and the verification image may be input into the comparison model, and based on the output of the comparison model, the difference between the reference image and the verification image may be determined. For example, as shown in FIG. 5 , the reference image 510 and the verification image 515 may be input into the comparison model 520 , and based on the output of the comparison model 520 , a difference 530 between the reference image 510 and the verification image 515 may be determined.

In some embodiments, the alignment model 520 may include a consistency check network. In some embodiments, the backbone network of the consistency checking network may include a classification structure (eg, a Resnet family of classification structures).

In some embodiments, the alignment model 520 may include a transformer structure. For more details about the transformer structure, please refer to FIG. 7 and related descriptions, and details will not be repeated here.

In some embodiments, with reference to step 310, the respective hues of the at least two colored lights may be set randomly, and at least one of brightness or saturation may be a variable value. Correspondingly, for at least two target images respectively corresponding to at least two beams of color light, the hue, brightness and saturation have the same regularity. Taking FIG. 6 as an example, the target image S6 is the verification image, and the target images S1, S2, S3, S4, and S5 are the reference images. Suppose the HSV parameters of the six target images S1, S2, S3, S4, S5 and S6 are G1(0,1,0.7), G2(0.3,1,0.8), G3(0.6,1,1), G4( 0.3, 1, 1), G5 (0.9, 1, 0.7), Q1 (0.3, 1, 1), where the S value (saturation) is all 1 (that is, the saturation is a fixed value of 1), and the H value (hue ) and the V value (brightness) are variable values (that is, both hue and brightness are random values between 0-1). For convenience of description, the HSV parameter of the verification image is referred to as a "check code", and the HSV parameter of the reference image is referred to as a "reference code".

Further, the processing device 112 may use a consistency check network to compare the check code with the five reference codes respectively. Since the S value (saturation) is all 1, it is only necessary to compare the difference in the H value (hue) and the V value (brightness). As shown in Table 1 below, if the hue and brightness of the check code and the reference code are the same, the difference is 11; if the hue and brightness of the check code and the reference code are not the same, the difference is 00; If the code and the reference code have the same hue but different brightness, the difference is 10; if the check code and the reference code have different hues but the same brightness, the difference is 01.

Table 1 Comparison example of check code and reference code

色调\亮度Hue\Brightness	相同same	不相同Are not the same
相同same	1111	1010
不相同Are not the same	0101	0000

Correspondingly, the verification code Q1 is compared with the five reference codes G1, G2, G3, G4 and G5 respectively, and the corresponding differences of the five reference images are obtained as follows:

Q1:G1=(0.3,1):(0,0.7)=00

Q1:G2=(0.3,1):(0.3,0.8)=10

Q1:G3=(0.3,1):(0.6,1)=01

Q1:G4=(0.3,1):(0.3,1)=11

Q1:G5=(0.3,1):(0.9,0.7)=00

In some embodiments, the processing device 112 divides both the verification image and the reference image into multiple image blocks (for example, 4 image blocks), uses the image block as a processing unit, and determines the verification image and the reference image based on the transformer structure. The differences between the corresponding image blocks are further synthesized to determine the final difference between the verification image and the reference image.

During the image verification process, the degree of influence of light on different regions of the target may be different, resulting in certain differences in image parameters (for example, HSV parameters) in different regions of the captured image. Correspondingly, by using the transformer structure to divide the image into multiple image blocks and then perform subsequent similarity processing and analysis, the refinement processing capability of the model can be improved, thereby improving the verification effect. For more details, please refer to FIG. 7 and its description.

In step 430, the target verification code is determined based on the differences corresponding to the reference images respectively. In some embodiments, step 410 may be accomplished by the target verification code determination module 230 .

In some embodiments, the target verification code determination module 230 may arrange and combine the differences corresponding to the respective reference images to determine the target verification code. In some embodiments, the target verification code determination module 230 may determine the target verification code by arranging the differences corresponding to the respective reference images in sequence in the order of the reference images. For example, in conjunction with the above, the target verification code may be 0010011100. In some embodiments, the target verification code determination module 230 may arrange the differences corresponding to the respective reference images in any order to determine the target verification code.

In some embodiments, the target verification code determination module 230 may perform other forms of processing on the differences corresponding to the respective reference images to determine the target verification code. For example, vectors, determinants, matrices, etc., are not limited in this specification.

It should be noted that the above-described differences in hue and brightness between the reference image and the verification image are only exemplary, and differences in other dimensions may also be determined according to actual conditions of image parameters of the reference image and the verification image. For example, if the brightness is set to a constant value, the difference in hue and saturation between the reference image and the verification image can also be determined. For another example, if hue, saturation, and hue are all changed values, the difference in three dimensions between the reference image and the verification image can be determined. In addition, in the above example, the combination of "0" and "1" is used to reflect the difference between the reference image and the verification image, and the difference between the two can also be expressed in other forms, such as letters, numbers, character strings, etc. This does not limit.

Figure 7 is a schematic diagram of an alignment model according to some embodiments of the present specification. As shown in Figure 7, the alignment model can be a transformer structure. Q represents the verification image, and its HSV parameter is (0.3, 1, 1); G represents the reference image, and its HSV parameter is (0.3, 1, 0.8).

The processing device 112 may divide the verification image Q into four image blocks of Q1, Q2, Q3, and Q4, and divide the reference image G into four image blocks of G1, G2, G3, and G4. Each image block is then input to a block encoding module (eg, the patch embedding module shown in FIG. 7 ), each image block is encoded into a token, and the token is input to an encoder module (eg, FIG. 7 ) transformer encoder shown in 7).

In some embodiments, the encoder module may be a two-head structure, eg, a first head shown on the left side of FIG. 7 and a second head shown on the right side.

In some embodiments, the encoder module may repeat the process multiple times (eg, L times) based on each marker code to determine the similarity relationship between the verification image and the reference image.

In some embodiments, in each processing, the encoder module may determine 4 sets of similarity matrices based on each marker code, each set includes two matrices H _QiGi and V _QiGi , which respectively represent image blocks corresponding to the two images Similarity in hue and brightness. In some embodiments, as shown in FIGS. 8A and 8B , the first head may generate the similarity matrix H _QiGi and the second head may generate the similarity matrix V _QiGi . For example, taking H _Q1G1 (representing the tonal similarity between the first image block of the verification image Q and the reference image G) as an example, H _Q1G1 includes four elements a _mn , where a ₁₁ =0.98, representing Q1 and Q1 Its own hue similarity; a ₂₂ =0.99, representing the hue similarity between G1 and G1; a ₁₂ =0.97, representing the hue similarity between Q1 and G1; a ₂₁ =0.99, representing the hue similarity between G1 and Q1.

Further, the sub-diagonal elements of the H _QiGi matrix corresponding to the multiple processing can be averaged to obtain the hue similarity score H* of the verification image Q and the reference image G, and the sub-diagonal of the V _QiGi matrix corresponding to the multiple processing can be obtained. The elements are averaged to obtain the brightness similarity score V* of the verification image Q and the reference image G. Then compare the hue similarity score H* and brightness similarity score V* with preset thresholds (for example, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, etc., which can be set based on actual needs), which are greater than the preset thresholds Then, it is considered that the image parameters (eg, hue, brightness) of the verification image Q and the reference image G are "consistent", and the corresponding difference is marked as 1, otherwise, it is marked as 0.

In some embodiments, the encoder module can further regress the global image parameters (for example, H value and V value). In some embodiments, as shown in FIG. 9 , the encoder module may further include an attention module, and through the attention module, spatial attention is constructed in the verification image Q and the reference image G respectively. By constructing spatial attention, the correlation between image blocks in the image can be considered, and the accuracy of the global spatial relationship can be improved.

In some embodiments, in the training process of the transformer structure, its loss function can be composed of two parts, namely the color parameter loss L1 and the similarity loss L2, wherein the color parameter loss L1 is a global image obtained by regressing two heads The parameters and their true labels are calculated by the mean square error, and the similarity loss L2 is obtained by calculating the cross-entropy loss of the similarity matrices H* and V* respectively generated in multiple processing processes and the corresponding label matrices H _target and V _target . An example of the label matrix is shown in Table 2 below, a _ij represents the color parameter similarity between the i-th image and the j-th image, and the elements a ₁₁ and a ₂₂ on the main diagonal of the matrix both represent the image itself and its own color parameters. Similarity (for example, hue similarity, brightness similarity), so the value is 1, and the elements on the sub-diagonal line represent the similarity of color parameters between different images, if the color parameters are the same, the value is 1, otherwise The value is 0.

Table 2 Label matrix example

a ₁₁ a ₁₁	a ₁₂ a ₁₂
a ₂₁ a ₂₁	a ₂₂ a ₂₂

Through the mutual promotion between the two training tasks, the expressive ability of the model can be improved, thereby improving the verification accuracy of the model when it is used. Specifically, in the training process, each image block contains its own local spatial information, and these local spatial relationships are interactively fused through the attention mechanism of the transformer, so as to obtain a global spatial relationship with a larger receptive field. This global spatial relationship can be Aided model training can extract richer and more effective image features, thereby improving the accuracy of similarity relationships.

It should be noted that the above description of the transformer structure is only exemplary and does not constitute a limitation. For example, the number of divisions of image blocks is not limited to 4 blocks, and may also be other numbers, such as 16 blocks, 64 blocks, etc., which is not limited in this specification.

In the embodiment of this specification, colored light is emitted during the process of collecting images, and the target verification code is determined based on the image parameters of the collected image. In addition, the reference verification code is determined based on the light parameters of the colored light, and the target verification code and the reference verification code are further compared. code for image authenticity verification, which can solve the problem of camera hijacking in verification scenarios. It is understandable that when illegal user B needs to verify, if illegal user B tries to obtain the verification image or video of legal user A through camera hijacking and verify, in this case, unless the parameters of legal user A's verification image or video It is completely consistent with the light parameters of the colored light emitted by the current system, otherwise it is impossible to pass the verification.

In addition, due to environmental conditions (for example, low ambient brightness, the presence of colored light in the environment, etc.) that will affect the image acquisition process, the acquired image may not accurately match the colored light emitted during the acquisition process, thereby affecting verification. result. In the embodiment of this specification, the target image is divided into a reference image and a verification image, and the target verification code is determined based on the difference between them (that is, the verification code is determined by self-comparison or self-inquiry); It is divided into reference light and verification light, and the reference verification code is determined based on the difference between them, and then the verification is carried out by comparing whether the target verification code and the reference verification code are consistent, which can eliminate the influence of environmental conditions and improve the accuracy of the verification result.

Further, in this specification, in addition to the color (hue) parameters of light or images, brightness and/or saturation are also considered. Accordingly, the number of digits of the verification code is more, which greatly reduces the probability of being hacked. Verification is also more accurate and secure. For example, assuming that the number of reference rays is N, the number of verification codes is 2N, and the probability of being illegally broken is 1/2 ^2N .

Furthermore, during the image verification process, the influence of light on different regions of the target may be different, resulting in certain differences in image parameters (eg, HSV parameters) in different regions of the captured image, which in turn affects the verification results. In the embodiment of this specification, the reference image and the verification image are respectively divided into multiple image blocks, and the subsequent similarity processing and analysis are performed based on the multiple image blocks through the transformer structure, which can improve the model refinement processing capability, thereby improving the verification. Effect.

FIG. 10 is a flowchart of another exemplary image verification process according to some embodiments of the present specification. As shown in Figure 10, the process 1000 includes the following steps:

Step 1010, acquiring multiple target images. The shooting time of the multiple target images has a corresponding relationship with the irradiation time of the multiple illuminations in the illumination sequence in which the terminal illuminates the target object.

In some embodiments, step 1010 may be performed by a target image acquisition module.

The target object refers to the object that needs to be recognized. For example, the target object may be a specific body part of the user, such as face, fingerprint, palm print, or pupil. In some embodiments, the target object refers to the face of a user requiring authentication and/or authorization. For example, in the online car-hailing application scenario, the platform needs to verify whether the driver who takes the order is a registered driver user reviewed by the platform, and the target object is the driver's face. For another example, in the face payment application scenario, the payment system needs to verify the payment authority of the payer, and the target object is the payer's face.

For target recognition of the target object, the terminal is instructed to emit a sequence of illuminations. A lighting sequence includes multiple lights that illuminate the target object. The colors of different lights in a lighting sequence can be the same or different. In some embodiments, the plurality of lights comprise at least two lights of different colors, ie the plurality of lights have multiple colors.

In some embodiments, the plurality of colors includes at least one reference color and at least one verification color. The verification color is one of the colors that is directly used to verify the authenticity of the image. The reference color is one of the colors used to assist verification in determining the authenticity of the target image. For more details about the reference color and the verification color, please refer to FIG. 3 and its related description, which will not be repeated here.

The lighting sequence contains information about each of the multiple lights, such as color information, lighting time, and so on. Color information for multiple lights in a lighting sequence can be represented in the same or different ways. For example, color information for multiple lights can be represented by color categories. For example, the colors of the multiple lights in the lighting sequence can be represented as red, yellow, green, purple, cyan, blue, and red. For another example, the color information of multiple lights can be represented by color parameters. For example, the colors of multiple lights in a lighting sequence can be represented as RGB(255, 0, 0), RGB(255, 255, 0), RGB(0, 255, 0), RGB(255, 0, 255), RGB (0, 255, 255), RGB(0, 0, 255). In some embodiments, a lighting sequence may also be referred to as a color sequence, which contains color information for multiple lights.

The illumination times of the plurality of illuminations in the illumination sequence may include the start time, end time, duration, etc., or any combination thereof, for each illumination plan to illuminate the target object. For example, the start time of illuminating the target object with red light is 14:00, and the start time of illuminating the target object with green light is 14:02. For another example, the durations for which the red light and the green light illuminate the target object are both 0.1 seconds. In some embodiments, the durations for different illuminations to illuminate the target object may be the same or different. The irradiation time can be expressed in other ways, which will not be repeated here.

In some embodiments, the terminal may sequentially emit multiple illuminations in a particular order. In some embodiments, the terminal may emit light through the light emitting element. The light-emitting element may include a light-emitting element built in the terminal, for example, a screen, an LED light, and the like. The light-emitting element may also include an externally-connected light-emitting element. For example, external LED lights, light-emitting diodes, etc. In some embodiments, when the terminal is hijacked or attacked, the terminal may receive an instruction to emit light, but does not actually emit light. For more details about the lighting sequence, please refer to FIG. 3 and its related description, which will not be repeated here.

In some embodiments, a terminal or processing device (eg, a target image acquisition module) may generate a lighting sequence randomly or based on preset rules. For example, a terminal or processing device may randomly select a plurality of colors from a color library to generate a lighting sequence. In some embodiments, the lighting sequence may be set by the user at the terminal, determined according to the default settings of the target recognition system 100, or determined by the processing device through data analysis, and the like. In some embodiments, the terminal or storage device may store the lighting sequence. Correspondingly, the target image acquisition module can acquire the illumination sequence from the terminal or the storage device through the network.

The multiple target images are images used for target recognition. The formats of multiple target images can include Joint Photographic Experts Group (JPEG), Tagged Image File Format (TIFF), Graphics Interchange Format (GIF), Kodak Flash PiX (FPX), Digital Imaging and Communications in Medicine (DICOM), etc. The multiple target images may be two-dimensional (2D, two-dimensional) images or three-dimensional (3D, three-dimensional) images.

In some embodiments, the target image acquisition module may acquire multiple target images. For example, the target image acquisition module may send acquisition instructions to the terminal through the network, and then receive multiple target images sent by the terminal through the network. Alternatively, the terminal may send multiple target images to a storage device for storage, and the target image acquisition module may acquire multiple target images from the storage device. The target image may not contain or contain the target.

The target image may be captured by the image acquisition device of the terminal, or may be determined based on data (eg, video or image) uploaded by the user. For example, in the process of target object verification, the target recognition system 100 will issue a lighting sequence to the terminal. When the terminal is not hijacked or attacked, the terminal can sequentially emit multiple lights according to the lighting sequence. When the terminal emits one of the multiple illuminations, its image acquisition device may be instructed to acquire one or more images within the illumination time of the illumination. Alternatively, the image capture device of the terminal may be instructed to capture video during the entire illumination period of the plurality of illuminations. The terminal or other computing device (eg, the processing device 112 ) may intercept one or more images collected during the illumination time of each illumination from the video according to the illumination time of each illumination. One or more images collected by the terminal during the irradiation time of each illumination can be used as multiple target images. At this time, the multiple target images are real images captured by the target object when the target object is illuminated by multiple lights. It can be understood that there is a corresponding relationship between the irradiation time of the multiple lights and the shooting time of the multiple target images. If one image is collected within the irradiation time of a single light, the corresponding relationship is one-to-one; if multiple images are collected within the irradiation time of a single light, the corresponding relationship is one-to-many.

When the terminal is hijacked, the hijacker can upload images or videos through the terminal device. Uploaded images or videos may contain target objects or specific body parts of other users, and/or other objects. The uploaded images or videos may be historical images or videos captured by the terminal or other terminals, or synthesized images or videos. A terminal or other computing device (eg, processing device 112) may determine a plurality of target images based on the uploaded image or video. For example, the hijacked terminal may extract one or more images corresponding to each illumination from the uploaded image or video according to the illumination sequence and/or illumination duration of each illumination in the illumination sequence. Just as an example, the lighting sequence contains five lights arranged in sequence, and the hijacker can upload five images through the terminal device. The terminal or other computing device will determine the image corresponding to each of the five lights according to the sequence in which the five images are uploaded. For another example, the irradiation time of the five lights in the lighting sequence is 0.5 seconds, respectively, and the hijacker can upload a 2.5-second video through the terminal. The terminal or other computing device can divide the uploaded video into five videos of 0-0.5 seconds, 0.5-1 seconds, 1-1.5 seconds, 1.5-2 seconds and 2-2.5 seconds, and take a screenshot of each video image. The five images captured from the video correspond to the five illuminations in sequence. At this point, the multiple images are fake images uploaded by the hijacker, not real images of the target object when illuminated by multiple lights. In some embodiments, if the image is uploaded by the hijacker through the terminal, the uploading time of the image or the shooting time in the video may be regarded as the shooting time. It can be understood that when the terminal is hijacked, there is also a corresponding relationship between the irradiation time of multiple lights and the shooting time of multiple images.

As mentioned above, the multiple colors corresponding to the multiple lights in the lighting sequence include at least one reference color and at least one verification color. In some embodiments, one or more of the at least one reference color is the same as one or more of the at least one verification color. The multiple target images include at least one reference image and at least one verification image, each of the at least one reference image corresponds to one of the at least one reference color, and each of the at least one verification image corresponds to one of the at least one verification color. a correspondence of .

For each of the multiple images, the target image acquisition module may use the color of the light corresponding to the irradiation time and the image capture time in the light sequence as the color corresponding to the image. Specifically, if the irradiation time of the light corresponds to the shooting time of one or more images, the color of the light is used as the color corresponding to the one or more images. It can be understood that when the terminal is not hijacked or attacked, the colors corresponding to the multiple images should be the same as the multiple colors of the multiple lights in the lighting sequence. For example, the multiple colors of multiple lights in the lighting sequence are "red, yellow, blue, green, purple, and red". When the terminal is not hijacked or attacked, the colors corresponding to the multiple images obtained by the terminal should also be "red, yellow". , blue, green, purple, red". When the terminal is hijacked or attacked, the colors corresponding to multiple images and multiple colors of multiple lights in the lighting sequence may be different.

Step 1020: For each of the at least one reference image, determine a target verification code based on the first color relationship between the reference image and each verification image.

In some embodiments, step 1020 may be performed by a verification module, wherein determining the first color relationship between the reference image and each verification image may be performed by a first color relationship determination module.

The first color relationship between the reference image and the verification image refers to the relationship between the color of the light when the reference image is captured and the color of the light when the verification image is captured. The first color relationship includes the same, different, or similar, and the like. In some embodiments, the first color relationship may be represented numerically. For example, the same is represented by "1", and the difference is represented by "0".

In some embodiments, the at least one first color relationship determined based on the at least one reference image and the at least one verification image may be represented by a vector, and each element in the vector may represent one and at least one of the at least one reference image. A first color relationship between one of the verification images. For example, if the first color relationship of each of the 1 reference image and the 5 verification images is the same, different, the same, the same, and different, then the first color relationship of the 1 reference image and the 5 verification images can be represented by a vector (1,0,1,1,0) means.

In some embodiments, the verification module may determine a target verification code (also referred to as a "first verification code") based on the first color relationship, and the target verification code includes a plurality of sub-codes represented by numerical values. The subcode for each position in the target verification code may represent a first color relationship between one of the at least one reference image and one of the at least one verification image. For example, the first color relationship between the above-mentioned one reference image and five verification images can be represented by the target verification code 10110.

In some embodiments, the first color relationship determination module may extract the reference color feature of the reference image and the verification color feature of each verification image. The first color relationship determination module may further determine the first color relationship between the reference image and each verification image based on the reference color feature and the verification color feature.

The reference color feature refers to the color feature of the reference image. Verifying color features refers to verifying the color features of an image. The color feature of an image refers to information related to the color of the image. The color of the image includes the color of the light when the image is captured, the color of the subject in the image, the color of the background in the image, and the like. In some embodiments, the color features may include deep features and/or complex features extracted by a neural network.

Color features can be represented in a number of ways. In some embodiments, the color feature can be represented based on the color value of each pixel in the image in the color space. A color space is a mathematical model that describes color using a set of numerical values, each of which can represent the color value of a color feature on each color channel of the color space. In some embodiments, a color space may be represented as a vector space, each dimension of the vector space representing a color channel of the color space. Color features can be represented by vectors in this vector space. In some embodiments, the color space may include, but is not limited to, RGB color space, Lαβ color space, LMS color space, HSV color space, YCrCb color space, HSL color space, and the like. Understandably, different color spaces contain different color channels. For example, the RGB color space includes red channel R, green channel G, and blue channel B, and color features can be represented by the color values of each pixel in the image on the red channel R, green channel G, and blue channel B, respectively.

In some embodiments, color features may be represented by other means (eg, color histograms, color moments, color sets, etc.). For example, the histogram statistics are performed on the color values of each pixel in the image in the color space to generate a histogram representing the color features. For another example, a specific operation (eg, mean, squared difference, etc.) is performed on the color value of each pixel in the image in the color space, and the result of the specific operation represents the color feature of the image.

In some embodiments, the first color relationship determination module may extract color features of the plurality of target images through a color feature extraction algorithm and/or a color verification model (or a portion thereof). Color feature extraction algorithms include color histogram, color moment, color set, etc. For example, the first color relationship determination module may count the gradient histogram based on the color value of each pixel in the image in each color channel of the color space, so as to obtain the color histogram. For another example, the first color relationship determination module may divide the image into multiple regions, and use the set of binary indices of the multiple regions established by the color values of each pixel in the image in each color channel of the color space to determine the image. color set.

In some embodiments, the first color relationship determination module may determine the similarity between the reference color feature of the reference image and the verification color feature of the verification image, and determine at least one first color relationship based on the similarity and the threshold. For example, if the similarity is greater than the first threshold, it is determined to be the same, if it is smaller than the second threshold, it is determined to be different, or larger than the third threshold and smaller than the first threshold, it is determined to be similar, and so on. The first threshold may be greater than the second threshold and the third threshold, and the third threshold may be greater than the second threshold. In some embodiments, the similarity may be characterized by the distance between the reference color feature and the verification color feature. The distance may include, but is not limited to, Euclidean distance, Manhattan distance, Chebyshev distance, Minkowski distance, Mahalanobis distance, included angle cosine distance, and the like.

In some embodiments, the first color relationship determination module may further acquire the first color relationship based on a color relationship determination layer included in the color verification model. For a detailed description of the color relationship determination layer, reference may be made to FIG. 12 and its related descriptions, which will not be repeated here.

Step 1030, for each of the at least one reference color, determine a reference verification code based on the reference color and the second color relationship of each verification color.

In some embodiments, step 230 may be performed by a verification module, wherein determining the reference color and the second color relationship for each verification color may be performed by a second color relationship determination module.

The second color relationship of the reference color and the verification color may indicate whether the two colors are the same, different, or similar. In some embodiments, the representation manner of the second color relationship may be similar to that of the first color relationship, and details are not described herein again. In some embodiments, the form of the reference verification code (which may also be referred to as a "second verification code") may be similar to the target verification code, which will not be repeated here.

In some embodiments, the second color relationship determination module may determine its second color relationship based on the reference color and the verification color category or color parameter. For example, if the categories in the reference color and the verification color are the same or the numerical difference of the color parameters is less than a certain threshold, the two colors are judged to be the same, otherwise, the two colors are judged to be different.

In some embodiments, the second color relationship determination module may extract the first color feature of the color template image of the reference color and the second color feature of the color template image of the verification color. The second color relationship determination module may further determine a second color relationship between the reference color and the verification color based on the first color feature and the second color feature. For example, the second color relationship determination module may calculate the similarity between the first color feature and the second color feature to determine the second color relationship.

In some embodiments, there is a one-to-one correspondence between at least one first color relationship and at least one second color relationship. Specifically, the first color relationship between the reference image and the verification image corresponds to the second color relationship between the reference color corresponding to the reference image and the verification color corresponding to the verification image.

Step 1040, based on the target verification code and the reference verification code, perform image authenticity verification.

In some embodiments, step 1040 may be performed by a verification module.

In some embodiments, image authenticity verification may be performed by determining the authenticity of multiple target images. The authenticity of the multiple target images can reflect whether the multiple target images are images captured by the target object under illumination of multiple colors. For example, when the terminal is not hijacked or attacked, its light-emitting element can emit light of multiple colors, and its image acquisition device can record or take pictures of the target object to obtain the target image. At this point, the target image is realistic. For another example, when the terminal is hijacked or attacked, the target image is obtained based on the image or video uploaded by the attacker. At this time, the target image does not have authenticity.

The authenticity of the target image can be used to determine whether the terminal's image capture device has been hijacked by an attacker. For example, if at least one target image in the multiple target images is not authentic, it means that the image acquisition device is hijacked. For another example, if more than a preset number of target images in the multiple target images are not authentic, it means that the image acquisition device is hijacked.

In some embodiments, the verification module may select part or all of at least one first color relationship to construct a corresponding first verification code, and construct a corresponding second verification code based on the second color relationship corresponding to the selected first color relationship , to determine the authenticity of multiple target images. Similar to the first vector and the second vector, the positions of the sub-codes in the first verification code and the second verification code are determined based on the correspondence between the first color relationship and the second color relationship. For example, if the first verification code and the second verification code are different, the multiple target images do not have authenticity. For example, if the first verification code is 10110 and the second verification code is 10111, the multiple target images are not authentic. For another example, the verification module may determine the authenticity of the multiple target images based on the same number of sub-codes in the first verification code and the second verification code. For example, if the number of identical subcodes is greater than the fifth threshold, the authenticity of the multiple target images is determined, and if the number of identical subcodes is less than the sixth threshold, it is determined that the multiple target images are not authentic. Exemplarily, the fifth threshold is 3, the sixth threshold is 1, the first verification code is 10110, the second verification code is 10111, and the first, second, and third digits of the first verification code and the second verification code are If the correspondence with the subcode of the fourth bit is the same, it is determined that the multiple target images are authentic.

In some embodiments, the verification module may determine the authenticity of the plurality of target images based on some or all of the at least one first color relationship and the corresponding second color relationship.

In some embodiments, the first color relationship and the second color relationship may be represented by vectors. In some embodiments, the verification module may select part or all of the at least one first color relationship to construct the first vector, and construct the second vector based on the second color relationship corresponding to the selected first color relationship. Further, the verification module may determine the authenticity of the multiple target images based on the similarity between the first vector and the second vector. For example, if the similarity is greater than the fourth threshold, the multiple target images are authentic. It can be understood that the arrangement order of elements in the first vector and the second vector is determined based on the corresponding relationship between the first color relationship and the second color relationship. For example, the element corresponding to a certain first color relationship in the first vector A is A _ij , and the element corresponding to the second color relationship corresponding to the first color relationship in the second vector B is B _ij .

As mentioned above, both the reference image and the verification image are captured under the same ambient light conditions and illuminated by the same light-emitting element. Therefore, based on the relationship between the reference image and the verification image, the In the case of authenticity, the influence of external ambient light and light-emitting elements can be eliminated or weakened, thereby improving the accuracy of light color recognition.

In some embodiments, the preset thresholds (eg, the fifth threshold, the sixth threshold) set for the image authenticity determination in some embodiments of this specification may be related to the degree of shooting stability. The shooting stability degree is the stability degree when the image acquisition device of the terminal acquires the target image. In some embodiments, the preset threshold is positively related to the degree of shooting stability. It can be understood that the higher the shooting stability, the higher the quality of the acquired target image, and the more the color features extracted based on multiple target images can truly reflect the color of the illumination when shooting, and the larger the preset threshold is. In some embodiments, the shooting stability may be measured based on a motion parameter of the terminal detected by a motion sensor of the terminal (eg, a vehicle-mounted terminal or a user terminal, etc.). For example, the motion speed, vibration frequency, etc. detected by the motion sensor. For example, the larger the motion parameter, or the larger the change rate of the motion parameter, the lower the shooting stability. The motion sensor may be a sensor that detects the driving situation of the vehicle, and the vehicle may be the vehicle used by the target user. The target user refers to the user to which the target object belongs. For example, if the target user is an online car-hailing driver, the motion sensor may be a motion sensor on the driver's end or the in-vehicle terminal.

In some embodiments, the preset threshold may also be related to the shooting distance and the rotation angle. The shooting distance is the distance between the target object when the image acquisition device acquires the target image. The rotation angle is the angle between the front of the target object and the terminal screen when the image acquisition device acquires the target image. In some embodiments, both the shooting distance and the rotation angle are negatively correlated with the preset threshold. It can be understood that the shorter the shooting distance, the higher the quality of the obtained target image, and the more the color features extracted based on multiple target images can truly reflect the color of the light at the time of shooting, the larger the preset threshold is. The smaller the rotation angle, the higher the quality of the acquired target image, and similarly, the larger the preset threshold. In some embodiments, the shooting distance and rotation angle may be determined based on the target image through image recognition techniques.

In some embodiments, the verification module may perform specific operations (eg, average, standard deviation, etc.) on the shooting stability, shooting distance, and rotation angle of each target image, and based on the specific calculation, the shooting stability, shooting distance, and The shooting angle determines the preset threshold.

For example, the verification module acquiring the stability degree of the terminal when multiple target images are acquired includes acquiring the sub-stability degree of the terminal when each of the multiple target images is captured; and fusing the multiple sub-stability degrees to determine the stability degree.

For another example, obtaining the shooting distance between the target object and the terminal when the multiple target images are captured by the verification module includes: acquiring the sub-shooting distance between the target object and the terminal when each of the multiple target images is captured; fusing the multiple sub-shooting distances to determine the shooting distance.

For another example, obtaining the rotation angle of the target object relative to the terminal when the multiple target images are captured by the verification module includes acquiring the sub-rotation angle of the target object relative to the terminal when each of the multiple target images is captured; Fusion to determine the rotation angle.

In some embodiments of the present specification, the target system 100 will issue a lighting sequence to the terminal, and acquire from the terminal a target image corresponding to a plurality of lightings in the lighting sequence. By recognizing the color of the illumination when the target face image is captured, the processing device can determine whether the target image is an image captured under the illumination sequence of the target object, and further determine whether the terminal is hijacked or attacked. It is understandable that when an attacker does not know the lighting sequence, it is difficult for the color of the light to be the same as the color of multiple lights in the light sequence when the uploaded image or the image in the uploaded video is captured. Even if the kinds of colors are the same, the order of the positions of each color is difficult to be the same. The method disclosed in this specification can improve the difficulty of an attacker's attack and ensure the security of target identification.

Figure 11 is a schematic diagram of a lighting sequence according to some embodiments of the present specification.

In some embodiments, the plurality of colors of lighting in the lighting sequence may include at least one reference color and at least one verification color. The verification color is one of the colors that is directly used to verify the authenticity of the image. The reference color is a color among the colors used to assist the verification color to determine the authenticity of the target image. For example, the target image corresponding to the reference color (also referred to as the reference image) may determine the first color relationship based on the target image corresponding to the verification color (also referred to as the verification image). Further, the verification module may determine the authenticity of the plurality of target images based on the first color relationship. As shown in Figure 3, the illumination sequence e contains multiple benchmark colors of illumination "red light, green light, blue light", and multiple verification colors of illumination "yellow light, purple light... cyan light"; the illumination sequence f contains multiple Lighting of the reference color "red light, white light...blue light", and light of multiple verification colors "red light..green light".

In some embodiments, multiple colors exist for verification. Multiple verification colors can be identical. For example, the verification color can be red, red, red, red. Alternatively, multiple verification colors can be completely different. For example, the verification color can be red, yellow, blue, green, violet. Alternatively, the plurality of verification colors may be partially the same. For example, the verification color can be yellow, green, purple, yellow, red. Similar to the verification color, in some embodiments, there are multiple reference colors, and the multiple reference colors may be identical, completely different, or partially identical. In some embodiments, the verification color may contain only one color, such as green.

In some embodiments, the at least one reference color and the at least one verification color may be determined according to default settings of the object recognition system 100, manually set by the user, or determined by the object image acquisition module. For example, the target image acquisition module may randomly select the reference color and the verification color. Just as an example, the target image acquisition module may randomly select a part of the colors from the plurality of colors as at least one reference color, and the remaining colors as at least one verification color. In some embodiments, the target image acquisition module may determine at least one reference color and at least one verification color based on preset rules. The preset rules may be rules about verifying the relationship between colors, the relationship between reference colors, and/or the relationship between verifying colors and reference colors, and the like. For example, the preset rule is that the verification color can be generated by fusion based on the reference color, and so on.

In some embodiments, one or more of the at least one reference color is the same as one or more of the at least one verification color. The at least one reference color and the at least one verification color may be completely identical or partially identical. For example, a certain one of the at least one verification color may be the same as a certain one of the at least one reference color. It can be understood that the verification color can also be determined based on at least one reference color, that is, the specific reference color can be used as the verification color. As shown in Figure 11, in the illumination sequence f, multiple reference colors "red, white...blue" and multiple verification colors "red...green" all contain red.

In some embodiments, the at least one reference color and the at least one verification color may also have other relationships, which are not limited herein. For example, at least one reference color and at least one verification color are the same or different in color family. Exemplarily, at least one reference color belongs to a warm color system (eg, red, yellow, etc.), and at least one reference color belongs to a cool color system (eg, gray, etc.).

In some embodiments, in the lighting sequence, the lighting corresponding to the at least one reference color may be arranged in front of or behind the lighting corresponding to the at least one verification color. As shown in Fig. 3, in the illumination sequence e, illuminations of multiple reference colors "red light, green light, blue light" are arranged in front of illuminations of multiple verification colors "yellow light, purple light...cyan light". In the illumination sequence f, illuminations of multiple reference colors "red light, white light...blue light" are arranged behind multiple verification colors "red light...green light". In some embodiments, the illumination corresponding to the at least one reference color may also be arranged at intervals with the illumination corresponding to the at least one verification color, which is not limited herein.

FIG. 12 is a schematic diagram of a color verification model according to some embodiments of the present specification.

In some embodiments, the verification module may determine the authenticity of the plurality of target images based on the color verification model. As shown in FIG. 12 , the color verification model may include a color feature extraction layer 1230 and a color relationship determination layer 1260 . Color feature extraction layer 1230 and color relationship determination layer 1260 may be used to implement step 1020. Further, the verification module may determine the authenticity of the plurality of target images based on the first color relationship and the second color relationship.

In some embodiments, the at least one reference image and the at least one verification image may form one or more image pairs. Each image pair includes one of at least one reference image and one of at least one verification image. The color verification model may separately analyze one or more image pairs to determine a first color relationship between the reference image and the verification image in the image pair. For example, as shown in FIG. 12, at least one reference image includes "1220-1, 1220-2...1220-y", and at least one verification image includes "1210-1...1210-x". For the purpose of illustration, the image pair formed by the reference image 1220-y and the verification image 1210-1 is taken as an example to expand.

The color feature extraction layer 1230 may extract the reference color feature of the reference image 1220-y and the verification color feature of the verification image 1210-1. In some embodiments, the type of the color feature extraction layer 1230 may include a Convolutional Neural Networks (CNN) model such as ResNet, ResNeXt, SE-Net, DenseNet, MobileNet, ShuffleNet, RegNet, EfficientNet, or Inception, or a loop Neural network model.

The input to the color feature extraction layer 1230 may be an image pair (eg, a reference image 1220-y and a verification image 1210-1). For example, the reference image 1220-y and the verification image 1210-1 can be concatenated and input to the color feature extraction layer 1230. The output of the color feature extraction layer 1230 may be the color features of the image pair. For example, the output of the color feature extraction layer 1230 may be the reference color feature 1250-y of the reference image 1220-y and the verification color feature 1240-1 of the verification image 1210-1. For another example, the output of the color feature extraction layer 1230 may be the color feature after splicing the color feature 1240-1 of the verification image 1210-1 and the color feature 1250-y of the reference image 1220-y.

The color relationship determination layer 1260 is configured to determine the first color relationship of the image pair based on the color features of the image pair. For example, the verification module may input the reference color feature 1250-y of the reference image 1220-y and the verification color feature 1240-1 of the verification image 1210-1 into the color relationship determination layer 1260, which outputs the reference image 1220-y and Verify the first color relationship of image 1210-1.

In some embodiments, the verification module may input multiple image pairs consisting of at least one reference image and at least one verification image together into the color verification model. The color verification model can simultaneously output the first color relationship for each of multiple pairs of images. In some embodiments, the verification module may input a pair of image pairs into the color verification model. The color verification model may output a first color relationship for the pair of images.

In some embodiments, the color relationship determination layer 1260 may be a classification model, including but not limited to fully connected layers, deep neural networks, decision trees, and the like.

In some embodiments, the color verification model is a machine learning model with preset parameters. Preset parameters refer to the model parameters learned during the training of the machine learning model. Taking a neural network as an example, the model parameters include weight and bias. The preset parameters of the color verification model are determined during the training process. For example, the model acquisition module can train an initial color verification model based on multiple training samples with labels to obtain a color verification model.

Training samples include one or more sample image pairs with labels. Each sample image pair includes two target images of the sample target object taken under the same or different lights. The labels of the training samples can indicate whether the sample image pairs were captured with the same color of lighting.

In some embodiments, the model acquisition module may input the training samples into the initial color verification model, and update the parameters of the initial color feature extraction layer and the initial color relationship determination layer through training until the updated color verification model satisfies preset conditions. The updated color verification model may be designated as a preset parameter color verification model, in other words, the updated color verification model may be designated as a trained color verification model. The preset condition may be that the loss function of the updated color verification model is smaller than the threshold, converges, or the number of training iterations reaches the threshold.

In some embodiments, the model acquisition module can train the initial color feature extraction layer and the initial color relationship determination layer in the initial color verification model through an end-to-end training method. The end-to-end training method means that the training samples are input into the initial model, the loss value is determined based on the output of the initial model, and the initial model is updated based on the loss value. The initial model may contain multiple sub-models or modules that perform different data processing operations, which are treated as a whole during training and updated simultaneously. For example, in the training of the initial color verification model, at least one sample reference image and at least one verification image can be input into the initial color feature extraction layer, and a loss function can be established based on the output results and labels of the initial color relationship determination layer. The parameters of each initial layer in the initial color verification model are updated simultaneously.

In some embodiments, the color verification model may be pre-trained by the processing device or a third party and stored in the storage device, and the processing device may directly call the color verification model from the storage device.

To determine the authenticity of multiple target images based on the first color relationship and the second color relationship, it is possible to identify whether the type of illumination when the target image is photographed is consistent directly by comparing the color features without identifying the type of illumination when the target image was photographed. It is equivalent to converting the color recognition task into a binary classification task of judging whether the colors are the same. In some embodiments, a color verification model may be used to determine the first color relationship. The color relationship determination model of the color verification model may include only a small number of neurons (eg, two neurons) to make the judgment of whether the colors are the same. Compared with the color recognition network in the traditional method, the structure of the color verification model disclosed in this specification is simpler. The target object analysis based on the color verification model also requires relatively less computing resources (eg, computing space), thereby improving the efficiency of light color recognition. Meanwhile, the input of the model can be a target image corresponding to any color. Compared with other algorithms that need to limit the number of input color types, the embodiment of this specification has higher applicability. Moreover, using the color verification model can improve the reliability of the authenticity verification of the target image, reduce or remove the influence of the performance difference of the terminal equipment, and further determine the authenticity of the target image. It can be understood that there are certain differences in the hardware of different terminals. For example, the color light of the same color emitted by the terminal screens of different manufacturers may have differences in parameters such as saturation and brightness, resulting in a large intra-class gap of the same color. The training samples of the initial color verification model can be taken by terminals with different performances. The initial color verification model is learned in the training process, so that the trained color verification model can consider the terminal performance difference when judging the color of the target object, and more accurately determine the color of the target image. In addition, when the terminal is not hijacked, both the reference image and the verification image are taken under the same ambient light conditions. Therefore, when the reference image and the verification image are processed based on the color verification model to determine the authenticity of multiple target images, the influence of external ambient light can be eliminated or reduced.

The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is merely an example, and does not constitute a limitation of the present specification. Although not explicitly described herein, various modifications, improvements, and corrections to this specification may occur to those skilled in the art. Such modifications, improvements, and corrections are suggested in this specification, so such modifications, improvements, and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

Meanwhile, the present specification uses specific words to describe the embodiments of the present specification. Such as "one embodiment," "an embodiment," and/or "some embodiments" means a certain feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various places in this specification are not necessarily referring to the same embodiment . Furthermore, certain features, structures or characteristics of the one or more embodiments of this specification may be combined as appropriate.

Furthermore, unless explicitly stated in the claims, the order of processing elements and sequences described in this specification, the use of alphanumerics, or the use of other names is not intended to limit the order of the processes and methods of this specification. While the foregoing disclosure discusses by way of various examples some embodiments of the invention presently believed to be useful, it is to be understood that such details are for purposes of illustration only and that the appended claims are not limited to the disclosed embodiments, but rather The requirements are intended to cover all modifications and equivalent combinations that fall within the spirit and scope of the embodiments of this specification. For example, although the system components described above may be implemented by hardware devices, they may also be implemented by software-only solutions, such as installing the described systems on existing servers or mobile devices.

Similarly, it should be noted that, in order to simplify the expressions disclosed in this specification and thus help the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of this specification, various features may sometimes be combined into one embodiment, in the drawings or descriptions thereof. However, this method of disclosure does not imply that the subject matter of the description requires more features than are recited in the claims. Indeed, there are fewer features of an embodiment than all of the features of a single embodiment disclosed above.

Some examples use numbers to describe quantities of ingredients and attributes, it should be understood that such numbers used to describe the examples, in some examples, use the modifiers "about", "approximately" or "substantially" to retouch. Unless stated otherwise, "about", "approximately" or "substantially" means that a variation of ±20% is allowed for the stated number. Accordingly, in some embodiments, the numerical parameters set forth in the specification and claims are approximations that can vary depending upon the desired characteristics of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and use a general digit reservation method. Notwithstanding that the numerical fields and parameters used in some embodiments of this specification to confirm the breadth of their ranges are approximations, in specific embodiments such numerical values are set as precisely as practicable.

For each patent, patent application, patent application publication, and other material, such as article, book, specification, publication, document, etc., cited in this specification, the entire contents of which are hereby incorporated by reference into this specification are hereby incorporated by reference. Application history documents that are inconsistent with or conflict with the contents of this specification are excluded, as are documents (currently or hereafter appended to this specification) limiting the broadest scope of the claims of this specification. It should be noted that, if there is any inconsistency or conflict between the descriptions, definitions and/or use of terms in the accompanying materials of this specification and the contents of this specification, the descriptions, definitions and/or use of terms in this specification shall prevail .

Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations are also possible within the scope of this specification. Accordingly, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to those expressly introduced and described in this specification.

Claims

An image verification method, the method includes:

Instruct the terminal equipment to emit at least two colored light beams to the target;

In the process of emitting the at least two colored light beams, instructing the terminal device to collect multiple images of the target;

Selecting at least two target images respectively corresponding to the at least two colored rays from the plurality of images;

determining a target verification code based on the at least two target images; and

Image authenticity verification is performed by comparing the target verification code with the reference verification codes corresponding to the at least two beams of colored light.
The method of claim 1, the at least two colored light beams comprising a plurality of colored light beams, wherein the at least two colored light beams are of different colors.
The method according to claim 1, the light parameters of the at least two colored lights include hue, saturation and brightness, wherein the hue is randomly set, and at least one of the saturation or the brightness is change value.
The method according to claim 1, wherein the determining of the target verification code based on the at least two target images comprises:

One of the at least two target images is used as a verification image, and the other is used as a reference image;

For each of the reference images, determining the difference between the reference image and the verification image; and

The target verification code is determined based on the differences corresponding to the reference images respectively.
5. The method of claim 4, wherein the difference represents at least one of a hue difference and a saturation difference or a brightness difference.
The method of claim 4, wherein the determining the difference between the reference image and the verification image comprises:

dividing the reference image and the verification image into a plurality of image blocks, respectively; and

Based on the plurality of image blocks, the difference between the reference image and the verification image is determined by a comparison model.
The method of claim 1,

The at least two colored light beams correspond to at least two colors respectively, and the at least two colors include at least one reference color and at least one verification color; and

The at least two target images include at least one calibration image and at least one reference image, each of the at least one reference image corresponds to one of the at least one reference color, the at least one calibration image. Each of the verification images corresponds to one of the at least one verification color.
8. The method of claim 7, wherein one or more of the at least one reference color is the same as one or more of the at least one verification color.
The method according to claim 7, wherein the performing image authenticity verification by comparing the target verification code with the reference verification codes corresponding to the at least two colored rays comprises:

For each of the at least one reference image, determine the target verification code based on the first color relationship between the reference image and each verification image;

for each of the at least one reference color, determining the reference verification code based on the reference color and a second color relationship of each of the verification colors; and

Based on the target verification code and the reference verification code, image authenticity verification is performed.
The method of claim 9,

each of the at least one reference image and each of the at least one verification image form at least one pair of images; and

For each of the at least one reference image, determining the first color relationship between the reference image and each of the verification images includes:

For each of the at least one pair of image pairs, the image pair is processed based on a color verification model to determine a first color relationship between the reference image and the verification image in the image pair, and the color verification model is A machine learning model with preset parameters.
The method according to claim 10, wherein the color verification model comprises a color feature extraction layer and a color relationship determination layer, wherein:

the color feature extraction layer extracts color features of the image pair; and

The color relationship determination layer determines a first color relationship between the reference image and the verification image in the image pair based on the color feature of the image pair.
The method according to claim 11, wherein the preset parameters of the color verification model are obtained through an end-to-end training method.
The method of claim 10, wherein the preset parameters of the color verification model are generated through a training process, the training process comprising:

Acquiring a plurality of training samples, each of the plurality of training samples includes a sample image pair and a sample label, the sample label indicating whether the sample images in the sample image pair are photographed under the illumination of the same color of light ;as well as

An initial color verification model is trained based on the plurality of training samples, and the preset parameters of the color verification model are determined.
The method of claim 9, wherein, for each of the at least one reference image, determining the first color relationship between the reference image and each of the verification images comprises:

extracting the reference color feature of the reference image and the verification color feature of each of the verification images; and

Based on the reference color feature and the verification color feature, a first color relationship between the reference image and each of the verification images is determined.
An image verification system, the system includes:

at least one storage medium storing computer instructions;

At least one processor, executing the computer instructions, implements the following methods:

Instruct the terminal equipment to emit at least two colored light beams to the target;

In the process of emitting the at least two colored light beams, instructing the terminal device to collect multiple images of the target;

Selecting at least two target images respectively corresponding to the at least two colored rays from the plurality of images;

determining a target verification code based on the at least two target images; and

Image authenticity verification is performed by comparing the target verification code with the reference verification codes corresponding to the at least two beams of colored light.
The system of claim 15, wherein the determining a target verification code based on the at least two target images comprises:

One of the at least two target images is used as a verification image, and the other is used as a reference image;

For each of the reference images, determining the difference between the reference image and the verification image; and

The target verification code is determined based on the differences corresponding to the reference images respectively.
The system of claim 16, the determining the difference between the reference image and the verification image comprising:

dividing the reference image and the verification image into a plurality of image blocks, respectively; and

Based on the plurality of image blocks, the difference between the reference image and the verification image is determined by a comparison model.
The system of claim 15,

The at least two colored light beams correspond to at least two colors respectively, and the at least two colors include at least one reference color and at least one verification color; and

The at least two target images include at least one calibration image and at least one reference image, each of the at least one reference image corresponds to one of the at least one reference color, the at least one calibration image. Each of the verification images corresponds to one of the at least one verification color.
The system according to claim 18, the performing image authenticity verification by comparing the target verification code with the reference verification codes corresponding to the at least two colored rays of light comprises:

For each of the at least one reference image, determine the target verification code based on the first color relationship between the reference image and each verification image;

for each of the at least one reference color, determining the reference verification code based on the reference color and a second color relationship of each verification color; and

Based on the target verification code and the reference verification code, image authenticity verification is performed.
A computer-readable storage medium storing computer instructions, when a computer reads the computer instructions, the computer executes the following method:

Instruct the terminal equipment to emit at least two colored light beams to the target;

In the process of emitting the at least two colored light beams, instructing the terminal device to collect multiple images of the target;

Selecting at least two target images respectively corresponding to the at least two colored rays from the plurality of images;

determining a target verification code based on the at least two target images;

Image authenticity verification is performed by comparing the target verification code with the reference verification codes corresponding to the at least two beams of colored light.