System suitable for detecting identity card and associated device and processing method

Abstract

A system suitable for detecting an identity card and an apparatus and processing method associated therewith are provided. The apparatus comprises: a photographing device for photographing data associated with an identity card associated with a customer, wherein the photographed data can be processed by a deep learning algorithm to obtain general prediction information about the identity card. The apparatus also comprises a processor for processing the general prediction information to specifically determine information about the identity card.

Description

Technical Field

[0001] The present invention generally, but not exclusively, relates to a system suitable for detecting identity cards and also relates to apparatus and / or processing methods associated with the system. Background Art

[0002] In the scenario of “electronic Know Your Customer / Client (eKYC)”, capturing data / information about the customer / client is crucial.

[0003] Traditionally, this can be accomplished, for example, by using a device to detect an identification card belonging to or associated with a customer / client. The identification card may include personal information / data (e.g., name, identification card number, nationality, and / or date of birth) associated with its owner (i.e., the customer / client to whom the identification card belongs or is associated).

[0004] One conventional solution / technique is to use an electronic device to scan an identification card belonging to or associated with a customer / client. Other conventional techniques include using an electronic device to capture an image of an identification card belonging to or associated with a customer / client.

[0005] However, it is contemplated herein that the data / information need not be captured in a reliable and / or efficient manner.

[0006] In particular, this document contemplates that conventional solutions / techniques do not facilitate detecting identification cards in a user-friendly, efficient and / or reliable manner.

[0007] This paper considers the need for improvements in the way identity cards can be detected. Summary of the invention

[0008] In order to improve the way identity cards can be detected, this paper considers a two-stage based detection strategy.

[0009] By considering the two-stage strategy-based approach, it is contemplated herein that identification cards can be detected in a more reliable / accurate manner without substantially relying on, for example, / over-reliance on, the optimal position of the identification card relative to the detection device. In addition, it may be possible to significantly reduce the number of attempts required to capture customer / client data / information (or, for example, only require one attempt). Therefore, it is further contemplated herein that identification cards can be detected in a user-friendly and / or efficient manner.

[0010] According to an embodiment of the present invention, a device is provided.

[0011] The apparatus comprises: a photographing device for photographing data associated with an identity card associated with a customer, wherein the photographed data can be processed by a deep learning algorithm to obtain general prediction information about the identity card. The apparatus also comprises a processor for processing the general prediction information to specifically determine information about the identity card.

[0012] According to another embodiment of the present invention, a processing method is provided.

[0013] The processing method may include a first-stage identity card detection step, in which general prediction information about the identity card is obtained based on a convolutional neural network (CNN).

[0014] The processing method may further include a second-stage identity card detection step, wherein information about the identity card is specifically determined based on the obtained general prediction information. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In the following, embodiments of the present invention are described with reference to the following drawings, in which:

[0016] Figure 1 An example of a system according to an embodiment of the present invention is shown, which system may include at least one device and / or at least one database;

[0017] Figure 2 The embodiment according to the present invention is shown in more detail Figure 1 Examples of devices;

[0018] Figure 3 The embodiment according to this invention is shown Figure 1 exemplary scenarios associated with the system;

[0019] Figure 4 A computer device according to an embodiment of the present invention is shown, which may correspond to Figure 1 Devices and / or Figure 1 or associated with a database;

[0020] Figure 5 The embodiment according to this invention is shown Figure 1 Examples of processing methods associated with the system; and

[0021] Figure 6 It is shown that according to the embodiments of this document, for example, Figure 2 An illustration of an example of modules of an apparatus of an apparatus. DETAILED DESCRIPTION

[0022] This article considers that, with respect to the conventional solutions / techniques discussed earlier, reliable detection of a customer / client's identification card using a detection device depends on the position of the identification card relative to the detection device.

[0023] For example, if the identification card is positioned relative to the detection device such that the captured image of the identification card is skewed (e.g., where the photo of the identification card is taken at a significant angle) and / or if the identification card is too far from the detection device, the data / information associated with the customer / client may not be accurately captured. This may result in erroneous data / information being captured and / or requiring multiple attempts to capture the customer / client data / information.

[0024] Therefore, it is contemplated herein that conventional solutions / techniques require that the identification card be optimally positioned relative to the detection device and, therefore, do not facilitate detection of the identification card in a user-friendly, efficient and / or reliable manner.

[0025] In order to improve the way that identity cards can be detected, this paper generally considers a two-stage detection, such as referring to Figures 1 to 6 discussed in further detail.

[0026] Embodiments will be described, by way of example only, with reference to the accompanying drawings, in which like reference numerals and characters indicate like elements or equivalents.

[0027] Some portions of the following description are presented, either explicitly or implicitly, in terms of algorithms and functional or symbolic representations of operations on data within a computer memory. These algorithmic descriptions and functional or symbolic representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. An algorithm is here, generally conceived to be a self-consistent sequence of steps leading to a desired result. These steps are steps requiring physical manipulation of physical quantities such as electrical, magnetic or optical signals capable of being stored, transferred, combined, compared, and otherwise manipulated.

[0028] Unless otherwise specifically stated, and as will become apparent from the following text, it will be understood that throughout this document, terms such as "receiving," "scanning," "calculating," "determining," "replacing," "generating," "initializing," "outputting," etc., refer to the actions and processes of a computer system or similar electronic device that manipulates and converts data represented by physical quantities within the computer system into other data represented by similar physical quantities within the computer system or other information storage, transmission, or display device.

[0029] One or more devices for performing the operation of one or more methods are also disclosed herein. Such devices may be specially constructed for the desired purpose, or may include a computer or other device selectively activated or reconfigured by a computer program stored in the computer. The algorithms and displays proposed herein are not inherently associated with any particular computer or other device. According to the teachings of this article, various machines may be used with the program. Alternatively, the construction of a more dedicated device for performing the desired method steps may be appropriate. The structure of a computer suitable for performing the various methods / processing described herein will emerge from the following description.

[0030] In addition, this paper also implicitly discloses a computer program, because it is clear to those skilled in the art that the various steps of the method described herein can be implemented by computer code. The computer program is not intended to be limited to any specific programming language and its implementation. It should be understood that various programming languages ​​and codes thereof can be used to implement the teachings of the disclosure contained herein. Moreover, the computer program is not intended to be limited to any specific control flow. Without departing from the spirit or scope of this paper, the computer program has many other variations, which can use different control flows.

[0031] In addition, one or more steps of the computer program can be performed in parallel rather than sequentially. Such a computer program can be stored on any computer readable medium. The computer readable medium can include a storage device such as a disk or optical disk, a memory chip, or other storage devices suitable for interfacing with a computer. The computer readable medium can also include a hard-wired medium such as illustrated in an Internet system, or a wireless medium such as illustrated in a GSM mobile phone system, and other wireless systems such as Bluetooth, ZigBee, Wi-Fi. When the computer program is loaded and executed on such a computer, the computer program effectively generates a device that realizes the steps of the preferred method.

[0032] In the following we will refer to Figures 1 to 6 The foregoing is discussed in more detail.

[0033] Reference Figure 1 , an example of a system 100 according to embodiments herein is shown. The system 100 may be adapted to detect at least one identification card that may belong to or be associated with at least one customer / client. The identification card may carry or be associated with information / data associated with the customer / client. According to embodiments herein, the detection of the identification card may be, for example, based on a two-stage based detection / strategy, as will be discussed in more detail later.

[0034] The system 100 may include one or more devices 102 , one or more databases 104 , and a communication network 106 .

[0035] The device 102 may be coupled to a database 104. Specifically, the device 102 may be coupled to the database 104 via a communication network 106, for example.

[0036] In one embodiment, the device 102 may be coupled to the communication network 106, and the database 104 may be coupled to the communication network 106. The coupling may be performed by wired coupling and / or wireless coupling. Generally, the device 102 may be configured to communicate with the database 104 via the communication network 106.

[0037] The apparatus 102 may correspond to, for example, one or more computers (eg, laptop computers, desktop computers, and / or electronic mobile devices having computing capabilities, such as smartphones and electronic tablets).

[0038] Typically, the device 102 may be configured to capture information / data associated with a customer / client. Based on the captured information / data (ie, information data associated with a customer / client), the device 102 may be configured to generate and transmit one or more input signals.

[0039] For example, the device 102 can be configured to capture at least one image associated with an identification card that carries information / data associated with a customer / client. It is contemplated herein that the captured image may contain false information / data (e.g., in addition to the image of the identification card, an unnecessary background image). In this regard, it is contemplated herein that the input signal may contain false data / information (i.e., unrelated to the data / information associated with the customer / client). It is contemplated herein that such false data / information may be reduced / eliminated.

[0040] According to embodiments of the present invention, an input signal may be transmitted from device 102 to database 104 for processing to generate one or more detection signals. The detection signal may, for example, correspond to information based on predictions about an identity card. The detection signal may be transmitted from database 104 to device 102 for processing. Device 102 may be configured to perform one or more processing tasks for the detection signal to generate one or more output signals. The output signal may, for example, correspond to an image in which false information / data of an identity card has been greatly reduced / substantially eliminated (e.g., unnecessary background images have been greatly reduced / substantially eliminated).

[0041] It is contemplated herein that the output signal may not be of the desired (or non-ideal) quality. For example, false data / information may not be reduced / eliminated in an appropriate manner, and / or data / information about the customer / client of the identification card may not be reproduced in an understandable / accurate manner (e.g., an exemplary case of skew).

[0042] According to an embodiment of the present invention, the device 102 can be configured to determine whether the output signal has an acceptable quality. If it is determined that the quality of the output signal is unacceptable / not enough / unsatisfactory, the device 102 can be further configured to process the detection signal by adjustment / refinement to generate one or more adjusted detection signals. The device 102 can also be further configured to transmit the adjusted detection signal to the database 104 for processing, thereby generating one or more improved detection signals. The improved detection signal can be transmitted from the database 104 to the device 102 for further processing, thereby generating an improved output signal.

[0043] The improved output signal can be considered to have a higher quality (e.g., more accurate, more understandable and / or more reduced false information / data) than the aforementioned output signal. Specifically, for example, the improved output signal can be considered to be a higher quality version (e.g., a more accurate version / more understandable version) of the output signal.

[0044] According to the embodiments of this article, reference will be made to Figure 2 The device 102 is discussed in more detail.

[0045] Database 104 may, for example, host a deep learning type algorithm / network, or may, for example, correspond to a deep learning type algorithm / network. In one example, database 104 may correspond to one or more host devices (e.g., one or more computers or a network of computers) hosting a convolutional neural network (CNN). In another example, database 104 may correspond to a CNN.

[0046] Typically, the database 104 may be configured to receive the input signal and / or the pre-processed input signal for further processing, so that a detection signal and / or an improved detection signal may be generated based on, for example, a CNN. The database 104 may be further configured to transmit the detection signal and / or the improved detection signal to the device 102 for further processing.

[0047] The communication network 106 may correspond to the Internet communication network, for example. The communication via the communication network 106 (ie, the communication between the device 102 and the database 104) may be performed by way of wired communication and / or wireless communication.

[0048] As mentioned before, this paper considers two-stage based detection.

[0049] The first stage detection may involve obtaining / deriving the aforementioned detection signal. For example, with respect to the first stage detection, a prediction-based process may be performed to generally predict information associated with the identity card, such as location (e.g., based on the aforementioned corner point(s) associated with the identity card portion 302).

[0050] The second stage detection may generally involve processing the detection signal to obtain / generate the aforementioned output signal. The second stage detection may optionally further involve generating / obtaining the aforementioned improved output signal based on the aforementioned improved detection signal. For example, with respect to the second stage detection, prediction-based processing may be performed to specifically (i.e., accurately) determine information associated with the identity card, such as a location (e.g., based on the aforementioned corner points associated with the processed identity card portion and / or the improved identity card portion 308 associated with the identity card).

[0051] Now, according to the embodiments of this invention, reference may be made to Figure 2 The device 102 is discussed in more detail.

[0052] Reference Figure 2 , according to embodiments herein, apparatus 102 is shown in more detail in the context of embodiment 200 .

[0053] In the embodiment 200, the apparatus 102 may include one or more photographing devices 202 and one or more processors 204. The photographing device 202 may be coupled to the processor 204. The coupling may be performed by wired coupling and / or wireless coupling.

[0054] The photographing device 202 may be configured to photograph information / data associated with a customer / client. In one example, the photographing device 202 may correspond to an image-capturing device (e.g., a camera) that may be configured to capture at least one image associated with an identity card, as described later in Figure 3 This will be discussed in further detail in the context of exemplary scenarios.

[0055] Based on the captured information / data (i.e., information / data associated with the customer / client), the capture device 202 can be configured to generate and transmit one or more input signals. The input signals can be transmitted from the capture device 202 to the processor 204 and / or the database 104. As previously described, the input signals can include false data / information (e.g., background images).

[0056] The processor 204 may be configured to receive an input signal. In addition, the processor 204 may be configured to transmit an input signal and / or process an input signal.

[0057] For example, the processor 204 may be configured to transmit the input signal to the database 104 for processing. The database 104 may be configured to process the input signal by a prediction-based process (e.g., CNN-based) to generate one or more detection signals. The detection signal may be transmitted to the processor 204 for further processing.

[0058] The processor 204 may be configured to perform one or more processing tasks on the detection signal.

[0059] In one embodiment, the processor 204 can be configured to perform transformation-based processing (e.g., perspective transformation-based processing). For example, the detection signal can be processed by perspective transformation-based processing to generate one or more output signals associated with the identity card. The output signal can, for example, correspond to an image of the identity card, and the above-mentioned false data / information can be greatly reduced / substantially eliminated. The processor 204 can be further configured to perform determination-based processing to determine whether the output signal has acceptable quality. For example, the determination of acceptable quality can be based on: whether false data / information can / has been reduced / eliminated in an appropriate manner, and / or whether data / information related to the customer / client of the identity card can / has been reproduced in an understandable / accurate manner.

[0060] In one embodiment, if the quality of the output signal is determined to be unacceptable / inadequate / unsatisfactory, the processor 204 may be further configured to perform adjustment / improvement based processing.

[0061] Specifically, if it is determined that the quality of the output signal is unacceptable / insufficient / unsatisfactory, the processor 204 can be further configured to process the received detection signal in a predetermined adjustment / improvement manner. Specifically, the received detection signal can be processed in a manner based on the adjustment / improvement processing to generate an adjusted detection signal.

[0062] The processing based on adjustment / improvement may, for example, involve the application of predetermined adjustment / improvement parameters. According to embodiments of the present invention, this will be referred to below. Figure 3 Further detailed discussion is provided for exemplary scenarios.

[0063] The adjusted detection signal can be transmitted from the device 102 (e.g., from the processor 204) to the database 104 for further processing. Based on the adjusted detection signal, the database 104 can be configured to generate one or more improved detection signals. The improved detection signal can be transmitted from the database 104 to the device 102 for processing (e.g., by the processor 204) to generate one or more improved output signals associated with the identity card. For example, the processor can receive and process the improved detection signal in a manner based on perspective transformation processing to generate an improved output signal associated with the identity card.

[0064] As mentioned above, the improved output signal can be considered to have a higher quality (e.g., more accurate / more understandable) than the above output signal. Specifically, for example, the improved output signal can be considered to be a higher quality version (e.g., more accurate / more understandable version) of the output signal.

[0065] In the following, reference will be made to Figure 3 The foregoing is discussed based on the exemplary scenario 300 .

[0066] refer to Figure 3 , showing an exemplary scenario 300 according to an embodiment of the present invention.

[0067] As mentioned above, according to embodiments of the present invention, the device 102 may be configured to, for example, capture at least one image associated with an identity card.

[0068] In the exemplary scenario 300, a captured image 300a, a processed image 300b, and an improved image 300c are discussed.

[0069] The captured image 300a may be associated with the identity card portion 302 and may include a background image portion 304. Specifically, the captured image 300a may correspond to an image that is a combination of the identity card portion 302 and the background image portion 304. The identity card portion 302 may be based on an "x" by "y" resolution (e.g., a 400 by 300 resolution; 400×300 resolution, where "x"=300 and "y"=400), and may be associated with one or more corner points. The background image portion 304 may be considered as false data / information. For example, in the captured image 300a, the proportion of the identity card portion 302 may be higher than that of the background image portion 304 (e.g., the ratio between the identity card portion 302 and the background image portion 304 is 80:20). In one example, approximately 70% of the captured image 300a may be the identity card portion 302, and approximately 30% of the captured image 300a may be the background image portion 304.

[0070] Additionally, according to embodiments herein, the identity card portion 302 associated with the captured image 300a may, for example, be visually perceived as being skewed (e.g., in the exemplary case where the identity card is poorly positioned and the image is captured / taken when the identity card is positioned at an extreme angle).

[0071] Furthermore, the aforementioned input signal may be associated with the photographed image 300a.

[0072] After the input signal (i.e., the signal associated with the captured image 300a) is transmitted from the device 102 and received by the database 104, the database 104 may process the captured image 300a by means of a predictive processing (e.g., based on the aforementioned CNN) to predict information related to the corner points, as represented by p1, p2, p3, and p4, of the identity card portion 302. The aforementioned detection signal may correspond to the predicted information about the corner points (i.e., p1, p2, p3, and p4).

[0073] The detection signal may be received by the device 102 for processing (eg, processing based on a perspective transformation) to generate the aforementioned output signal that may correspond to the processed image 300b.

[0074] The processed image 300b can be associated with the processed identity card portion 306. In addition, the background image portion 304 described above can be greatly reduced / substantially eliminated compared to each captured image 300a. For example, in the processed image 300b, the processed identity card portion 306 can be a much higher proportion than the background image portion 304 (e.g., the ratio between the processed identity card portion 306 and the background image portion 304 is 100:1). In one example, 90% or more of the processed image 300b can be the processed identity card portion 306, while 10% or less of the processed image 300b can be the background image portion 304. In addition, in one embodiment, the processed identity card portion 306 can be considered to be visually perceived as less skewed than the identity card portion 302. Specifically, the degree of skew associated with the processed identity card portion 306 can be less than that of the identity card portion 302. In one embodiment, skew can be substantially eliminated (eg, the processed identification card portion 306 can be substantially visually perceived as not skewed).

[0075] For example, the captured image 300a and the processed image 300b may each be associated with a ratio between identification (i.e., identification card portion 302 / processed identification card portion 306) and false data (i.e., background image portion 304). The ratio between identification and false data associated with the processed image 300b may be considered better / improved compared to the captured image 300a. Specifically, the ratio between the processed identification card portion 306 and the background image portion 304 may be higher for each processed image 300b than the ratio between the identification card portion 302 and the background image portion 304 for each captured image 300a. More specifically, the ratio of the processed identification card portion 306 (i.e., relative to the background image portion 304) for each processed image 300b may be considered much higher than the ratio of the identification card portion 302 (i.e., relative to the background image portion 304) for each captured image 300a.

[0076] In this regard, the processed image 300b can, for example, correspond to a set of output signals in which data / information associated with the customer / client (i.e., associated with the processed identification card portion 306) can be more accurately reproduced in a more understandable manner (i.e., compared to the customer or client data / information associated with the identification card portion 302).

[0077] The processed identity card portion 306 may be associated with one or more corner points. For example, the processed identity card portion 306 may be associated with four corner points as represented by p1', p2', p3', and p4'. Information related to the corner points of the processed identity card portion 306 may be derived / obtained / determined based on, for example, a perspective transform-based process for the corner points of the identity card portion 302.

[0078] For example, p1, p2, p3, and p4 may be processed by a perspective transformation based process to generate p1', p2', p3', and p4', respectively.

[0079] According to an embodiment of the present invention, in a context based on a 400 by 300 resolution according to the exemplary scene 300, p1' may be predefined as (0,0), p2' may be predefined as (400,0), p3' may be predefined as (400,300), and p4' may be predefined as (400,0).

[0080] Suitably, the processed image 300b may be viewed as an improved version of the captured image 300a, for example, because the data / information associated with the customer / client may be reproduced / obtained / determined in a more accurate / understandable manner. In this regard, it is expected that the identification card may be detected in a more reliable / accurate manner without substantially having to, for example, rely / over-rely on the optimal positioning of the identification card relative to the detection device. Furthermore, it may be possible to significantly reduce the number of attempts required to capture the customer / client data / information (or, for example, only one attempt may be required). Thus, the present invention may also, for example, facilitate the detection of identification cards in a user-friendly and / or efficient manner.

[0081] It is contemplated herein that further improvements may be needed / desired for the processed image 300b. For example, the processed image 300b may be deemed to still be of insufficient / unacceptable / undesirable quality.

[0082] Thus, according to embodiments herein, an improved image 300c may be generated that may be considered to be of higher quality (eg, more accurate / more understandable) than the processed image 300b. The improved image 300c may, for example, correspond to the aforementioned improved output signal.

[0083] Specifically, it is contemplated herein that, in one embodiment, an improved image 300c may be generated based on the above-described adjusted detection signal. The improved image 300c may be associated with or may substantially correspond to the improved identification card portion 308.

[0084] For example, one or more improvement / adjustment parameters may be applied to the received detection signal to generate an adjusted detection signal.Adjustments / improvements may be made, for example, related to the width and / or height associated with the identity card portion 302 associated with the captured image 300a.

[0085] Specifically, the identity card portion 302 may be associated with a width (represented by "w") and a height (represented by "h"). For example, with respect to the detection signal (i.e., as previously described, which may correspond to prediction information regarding corner points "p1", "p2", "p3", and "p4" associated with the identity card portion 302), a predetermined adjustment / improvement factor of the product of 0.1 and w (i.e., 0.1×w or 0.1*w) may be applied for "x" as follows, and a predetermined adjustment / improvement factor of the product of 0.1 and h (i.e., 0.1×h or 0.1*h) may be applied for "y" as follows:

[0086] p1(x,y)=p1(x-0.1*w,y-0.1*h)

[0087] p2(x,y)=p2(x+0.1*w,y-0.1*h)

[0088] p3(x,y)=p3(x+0.1*w,y+0.1*h)

[0089] p4(x,y)=p4(x-0.1*w,y+0.1*h)

[0090] In this regard, according to embodiments herein, it will be appreciated that the adjusted detection signal may include p1 (x, y), p2 (x, y), p3 (x, y), and / or p4 (x, y).

[0091] According to an embodiment of the present invention, in the context of a 400 by 300 resolution according to an exemplary scene 300, p1(x, y) may be predefined as (0, 0) (e.g., x=0; y=0), p2(x, y) may be predefined as (400, 0) (e.g., x=400; y=0), p3(x, y) may be predefined as (400, 300) (e.g., x=400; y=300), and p4(x, y) may be predefined as (400, 0) (e.g., x=400; y=0).

[0092] The adjusted detection signals (e.g., p1(x,y), p2(x,y), p3(x,y) and / or p4(x,y)) can be transmitted to the database 104 for processing (e.g., based on CNN) to generate improved prediction-based information (e.g., corner points) related to the identity card. The improved prediction-based information about, for example, the corner points of an identity card can be represented, for example, by p1", p2", p3", and p4". For example, p1(x,y), p2(x,y), p3(x,y) and / or p4(x,y) can be processed based on CNN to generate p1", p2", p3" and / or p4", respectively. In this regard, it can be understood that the improved detection signals can include p1", p2", p3" and / or p4".

[0093] For example, p1 ″, p2 ″, p3 ″ and / or p4 ″ may be processed by perspective transformation-based processing (eg, by device 102 ) to generate pp1 , pp2 , pp3 and / or pp4 , respectively.

[0094] The improved identity card portion 308 may be associated with one or more corner points, or may include one or more corner points, which may be represented by pp1, pp2, pp3, and / or pp4. According to embodiments herein, in the context of a 400 by 300 resolution according to the exemplary scene 300, pp1 may be predefined as (0,0), pp2 may be predefined as (400,0), pp3 may be predefined as (400,300), and pp4 may be predefined as (400,0).

[0095] In one embodiment, the improved image 300c should ideally be free of the background image portion 304. In this regard, the improved image 300c may, for example, substantially correspond to the improved identification card portion 308 (i.e., the background image portion 304 is absent / the background image portion 304 is considered negligible (if any)). Thus, in the improved image 300c, the above-mentioned background image portion 304 (the background image portion of each captured image 300a and / or processed image 300b) may be considered to have been at least substantially eliminated (or negligible). For example, more than 98% (e.g., ideally 100%) of the improved image 100c may be the improved identification card portion 308, while 2% or less (e.g., ideally 0%) of the improved image 300c may be the background image portion 304. Therefore, in the improved image 300c, the proportion of the improved ID card portion 308 compared to the background image portion 304 can be considered to be much higher (for example, the ratio between the improved ID card portion 308 and the background image portion 304 is 100:0.1 or 0) compared to the processed image 300b.

[0096] Furthermore, in one embodiment, for example, the improved identity card portion 308 may be considered to be visually perceived as having less skew than the processed identity card portion 306. Specifically, the degree of skew associated with the improved identity card portion 308 may be less than that of the processed identity card portion 306. In one embodiment, the skew may be substantially eliminated (i.e., the improved card portion 308 is visually perceived as being non-skewed).

[0097] As previously described, the improved output signal (i.e., corresponding to the improved image 300c) can be considered to have a higher quality (e.g., more accurate / more understandable) than the aforementioned output signal (i.e., corresponding to the processed image 300b). Specifically, for example, the improved image 300c can be considered to be a higher quality version (e.g., a more accurate version / more understandable version) of the processed image 300b.

[0098] In this regard, it can be expected that one technical effect is that the identification card can be detected in a more reliable / accurate manner without, for example, relying / over-relying on the optimal positioning of the identification card relative to the detection device. For example, even in the exemplary case where the identification card is poorly positioned relative to, for example, the device 102 (for example, at an extreme angle), the identification card can be detected in a reliable / accurate manner. In addition, the number of attempts required to capture the customer / client data / information may be greatly reduced (or, for example, only one attempt may be required). Therefore, another technical effect may be that the detection of the identification card in a user-friendly manner and / or an efficient manner may be facilitated.

[0099] To implement the above content, according to an embodiment, a device 102 is considered herein, which may include a photographing device 202 and a processor 204. The photographing device 202 may be coupled to the processor 204, for example.

[0100] The capturing device 202 may be configured to capture data associated with an identification card that may be associated with a client or customer. The captured data (e.g., data associated with the aforementioned input signal that may correspond to the captured image 300a) may be processed by a deep learning algorithm to obtain general prediction information about the identification card (e.g., information corresponding to the aforementioned detection signal). The deep learning algorithm may be hosted, for example, by the database 104.

[0101] The processor 204 can be configured to process general prediction information (e.g., information corresponding to the above-mentioned detection signal) to specifically determine information about the identity card (e.g., determine / generate the aforementioned output information that may correspond to the processed image 300b and the aforementioned improved output signal that may correspond to the improved image 300c; input signal and / or improved output signal).

[0102] In one embodiment, the deep learning algorithm may correspond to a convolutional neural network (CNN), which may be configured to obtain / derive / determine general predictive information related to at least one corner point (e.g., p1, p2, p3, and / or p4) associated with the identity card.

[0103] In one embodiment, the processor 204 may be further configured to process the general prediction information (e.g., information corresponding to the aforementioned detection signal) to generate adjusted improved prediction information (e.g., information corresponding to the aforementioned adjusted detection signal), which may be processed by a deep learning algorithm (e.g., CNN) to derive improved prediction information (e.g., information corresponding to the aforementioned improved detection signal). For example, the information specifically determined about the identity card may be based on the improved prediction information (e.g., obtaining / deriving / determining the aforementioned improved output signal based on the aforementioned improved detection signal).

[0104] In one embodiment, the general prediction information may be related to a plurality of corner points (e.g., p1, p2, p3, and / or p4) associated with the identification card. The processor 204 may be configured to process the general prediction information by a perspective transformation-based processing method to specifically determine information about the identification card (e.g., information corresponding to the above-mentioned output signal that may correspond to the processed image 300b).

[0105] In one embodiment, perspective transform based processing may be performed in conjunction with general prediction information (eg, p1, p2, p3, and / or p4) to generate a plurality of defined corner points (eg, p1', p2', p3', and / or p4').

[0106] In one embodiment, the processor 204 may be further configured to process the general prediction information (e.g., p1 to p4) by adjustment-based processing to derive a plurality of adjusted corner points (e.g., corner points corresponding to the aforementioned adjusted detection signals, which may, for example, include p1(x, y), p2(x, y), p3(x, y) and / or p4(x, y)). For example, the adjustment-based processing may involve applying an adjustment factor to the general prediction information (e.g., p1 to p4).

[0107] Furthermore, in one embodiment, the ID card may be associated with a width (w) and a height (h). Additionally, the ID card may be associated with a resolution defined based on "x" times "y" (e.g., 400×300 resolution). The adjustment may be, for example, 0.1. The adjusted corners may be defined as follows:

[0108] p1(x,y)=p1(x-0.1*w,y-0.1*h)

[0109] p2(x,y)=p2(x+0.1*w,y-0.1*h)

[0110] p3(x,y)=p3(x+0.1*w,y+0.1*h)

[0111] p4(x,y)=p4(x-0.1*w,y+0.1*h)

[0112] In one embodiment, the processor 204 may be configured to transmit the adjusted corner points (e.g., p1(x,y), p2(x,y), p3(x,y) and / or p4(x,y)) for further processing, such as by a CNN, to obtain improved prediction information (e.g., p1", p2", p3" and / or p4"). The processor 204 may be further configured to process the improved prediction information by a perspective transform-based processing to derive a plurality of improved corner points (e.g., pp1, pp2, pp3 and / or pp4).

[0113] It is understandable that information about the identity card can be specifically determined based on defined corner points (eg, p1', p2', p3' and / or p4') and / or improved corner points (pp1, pp2, pp3 and / or pp4)).

[0114] It is contemplated herein that in one example, the apparatus 102 and / or the database 104 may correspond to / be associated with a computing device. Figure 4 This is discussed in the context of an exemplary computing device.

[0115] Figure 4 An exemplary computing device 400 is shown, hereinafter interchangeably referred to as computer system 400, in accordance with an embodiment of the present disclosure. The following description of computing device 400 is provided by way of example only and is not intended to be limiting.

[0116] like Figure 4 As shown, exemplary computing device 400 includes a processor 404 for executing software routines. Although a single processor is shown for clarity, computing device 400 may also include a multi-processor system. Processor 404 is connected to a communication infrastructure 406 to communicate with other components of computing device 400. Communication infrastructure 406 may include, for example, a communication bus, a crossbar switch, or a network.

[0117] The computing device 400 also includes a main memory 408, such as a random access memory (RAM), and a secondary memory 410. The secondary memory 410 may include, for example, a hard disk drive 412 and / or a removable storage drive 414, which may include a floppy disk drive, a tape drive, an optical disk drive, etc. The removable storage drive 414 reads and / or writes to a removable storage unit 418 in a well-known manner. The removable storage unit 418 may include a floppy disk, a tape, an optical disk, etc. that is read and written by the removable storage drive 414. As will be appreciated by those skilled in the relevant art, the removable storage unit 418 includes a computer-readable storage medium having computer-executable program code instructions and / or data stored therein.

[0118] In alternative embodiments, the secondary memory 410 may additionally or alternatively include other similar means for allowing computer programs or other instructions to be loaded into the computing device 400. Such means may include, for example, a removable storage unit 422 and an interface 420. Examples of removable storage units 422 and interfaces 420 include program cartridges and cartridge interfaces (such as those found in video game console devices), removable storage chips (such as EPROMs or PROMs) and associated slots, and other removable storage units 422 and interfaces 420 that allow software and data to be transferred from the removable storage unit 422 to the computer system 400.

[0119] The computing device 400 also includes at least one communication interface 424. The communication interface 424 allows software and data to be transmitted between the computing device 400 and an external device via a communication path 426. In various embodiments herein, the communication interface 424 allows data to be transmitted between the computing device 400 and a data communication network such as a public data or private data communication network. The communication interface 424 can be used to exchange data between different computing devices 400, which form a part of an interconnected computer network. Examples of the communication interface 424 can include a modem, a network interface (such as an Ethernet card), a communication port, an antenna with associated circuits, etc. The communication interface 424 can be wired or can be wireless. The software and data transmitted via the communication interface 424 are in the form of a signal, which can be an electrical, electromagnetic, optical or other signal that can be received by the communication interface 424. These signals are provided to the communication interface by a communication path 426.

[0120] like Figure 4 As shown, computing device 400 also includes a display interface 402 that performs operations for presenting images to an associated display 430 , and an audio interface 432 that performs operations for playing audio content via associated speakers 434 .

[0121] As used herein, the term "computer program product" may refer in part to removable storage unit 418, removable storage unit 422, a hard disk installed in hard disk drive 412, or a carrier wave that carries the software to communication interface 424 via communication path 426 (wireless link or cable).

[0122] Computer-readable storage media refers to any non-transitory tangible storage media that provides recorded instructions and / or data to the computing device 400 for execution and / or processing. Examples of such storage media include floppy disks, tapes, CD-ROMs, DVDs, Blu-ray TM ) optical disk, hard drive, ROM or integrated circuit, USB memory, magneto-optical disk, or computer readable card such as PCMCIA card, whether these devices are internal or external to computing device 400. Examples of transitory or non-tangible computer-readable transmission media that may also be involved in providing software, applications, instructions and / or data to computing device 400 include radio or infrared transmission channels and network connections to another computer or networked device, as well as the Internet or Ethernet, including email transmissions and information recorded on websites, etc.

[0123] Computer programs (also referred to as computer program code) are stored in the main memory 408 and / or the secondary memory 410. The computer programs may also be received via the communication interface 424. Such computer programs, when executed, enable the computing device 400 to perform one or more features of the embodiments discussed herein. In various embodiments, the computer programs, when executed, enable the processor 404 to perform the features of the above-described embodiments. Thus, such computer programs represent controllers of the computer system 400.

[0124] The software may be stored in a computer program product and may be loaded into computing device 400 using removable storage drive 414, hard drive 412, or interface 420. Alternatively, the computer program product may be downloaded to computer system 400 via communications path 426. The software, when executed by processor 404, causes computing device 400 to perform the functions of the embodiments described herein.

[0125] It should be understood that Figure 4 The embodiments are provided by way of example only. Thus, in some embodiments, one or more features of computing device 400 may be omitted. Moreover, in some embodiments, one or more features of computing device 400 may be combined together. Additionally, in some embodiments, one or more features of computing device 400 may be separated into one or more components.

[0126] It should be understood that Figure 4 The illustrated elements are used to provide means for performing various functions and operations of the server as described in the above embodiments.

[0127] In an embodiment, a server can generally be described as a physical device including at least one processor and at least one memory including computer program code. The at least one memory and the computer program code are configured to work with the at least one processor to enable the physical device to perform necessary operations.

[0128] Reference Figure 5 , showing an example of a processing method 500 associated with the system 100 according to an embodiment of the present invention.

[0129] The processing method 500 may include any one or any combination of an input step 502, a processing (or "generating") step 504, and an output step 506. According to an embodiment of the present invention, the processing method 500 may further include an improving step 508. According to an embodiment of the present invention, the processing method 500 may further include a training step 510.

[0130] According to embodiments herein, the processing method 500 may be associated with, for example, a two-stage detection of an identity card.

[0131] Regarding the input step 502, the aforementioned input signal may be generated. Specifically, one or more input signals associated with an identity card may be generated.

[0132] Regarding processing step 504, the input signal can be processed to determine / obtain prediction-based information about the identification card. The prediction-based information about the identification card can, for example, correspond to the aforementioned detection signal. In addition, according to embodiments of the present invention, the input signal can be processed via a deep learning algorithm. For example, the input signal can be processed based on a CNN.

[0133] Regarding the output step 506, the aforementioned output signals may be generated. Specifically, one or more output signals may be generated based on the detection signal.

[0134] With respect to the improving step 508, the aforementioned improved output signal may be generated. As previously described, one or more improved output signals may be generated based on one or more improved detection signals.

[0135] Regarding the training step 510, the deep learning algorithm can be trained. The training of the deep learning algorithm can be based on one or more output signals and / or improved one or more output signals.

[0136] As mentioned above, in one embodiment, the processing method 500 can be associated with a two-stage detection of an identity card. According to embodiments of this document, the two-stage detection of an identity card can be associated with, for example, a first-stage identity card detection step and a second-stage identity card detection step.

[0137] The first stage detection (of the identification card) may be associated with an input step 502 and a processing step 504 .

[0138] The second stage detection (of the identification card) can be associated with the output step 506 and / or the improvement step 508.

[0139] It is understandable that, in one embodiment, the training step 510 can be based on the first stage detection and / or the second stage detection. For example, the CNN can be trained based on the first stage detection and the second stage detection.

[0140] Generally, in one embodiment, contemplated herein is a processing method 500 that may include a first stage identity card detection step and a second stage identity card detection step.

[0141] Regarding the first stage identity card detection step, general prediction information about the identity card (e.g., p1, p2, p3 and / or p4) can be obtained based on, for example, a convolutional neural network (CNN).

[0142] Regarding the second stage identity card detection step, prediction information about the identity card can be determined based on the obtained general prediction information (e.g., p1, p2, p3 and / or p4) (e.g., by means of perspective transformation-based processing).

[0143] It is contemplated herein that in one embodiment, a CNN may be trained based on a first stage identity card detection step and / or a second stage identity card detection step.

[0144] refer to Figure 6 , shows an example of modules of an apparatus 600 according to an embodiment of this document. The apparatus 600 may be an example of an embodiment of the aforementioned apparatus 102.

[0145] As shown, the apparatus 600 may include a first module 602 and a second module 604. The first module 602 may be associated with the aforementioned first stage detection (of the identity card). The second module 604 may be associated with the aforementioned second stage detection (of the identity card).

[0146] Specifically, the apparatus 600 may include a first module 602 for acquiring / photographing data / information associated with a customer / client, and generating one or more input signals (e.g., a signal corresponding to the photographed image 300a) based on the photographed / acquired data / information. As described above, one or more detection signals may be generated based on the input signal.

[0147] The apparatus 600 may also include a second module 602 for processing the detection signal to generate one or more output signals (e.g., a signal corresponding to the processed image 300b). In one embodiment, the second module 602 may generate one or more improved output signals (e.g., a signal corresponding to the improved image 300c). In one embodiment, the detection signal may be further processed to obtain one or more improved detection signals. The one or more improved output signals may be generated based on the improved detection signal.

[0148] For example, Figure 6 It can be interpreted as illustrating the internal functional modules and / or structures of the detection device that can detect the identity card based on the two-stage detection strategy according to the embodiments of this document.

[0149] Those skilled in the art will further appreciate that variations and combinations of the above features, rather than substitutions or replacements, may be combined to form other embodiments.

[0150] In one example, the task of determining-based processing may be omitted. Specifically, by default (ie, without determining whether the output signal is of acceptable / sufficient quality), the processor may be configured to perform adjustment / improvement-based processing on the detection signal to generate one or more improved detection signals.

[0151] In another example, according to embodiments herein, the database 104 may be configured to generate one or more improved output signals based on a combination of the detection signal and the improved detection signal.

[0152] In the foregoing manner, various embodiments of the present invention are described to address at least one of the aforementioned disadvantages. Such embodiments are intended to be covered by the appended claims and are not limited to the specific forms or arrangements of the components described, and in view of this disclosure, it will be apparent to those skilled in the art that many changes and / or modifications may be made, which are also intended to be covered by the appended claims.

Claims

1. A device for detecting an identity card, comprising: A photographing device for photographing data associated with an identification card associated with a customer, wherein the photographed data can be processed by a deep learning algorithm to obtain general prediction information related to at least one corner point of the identification card; as well as a processor for processing the general prediction information to specifically determine information about the identity card; The processor is further configured to: process general prediction information related to the at least one corner point based on a perspective transformation to obtain a corresponding defined corner point, and adjust the at least one defined corner point according to an adjustment factor and a width and / or height associated with the identity card, and process the adjusted corner point based on a perspective transformation to obtain an improved corner point; the improved corner point is used to determine the information of the identity card.

2. The device according to claim 1, wherein: The deep learning algorithm corresponds to a convolutional neural network that derives general prediction information related to at least one corner point associated with the identification card.

3. According to the device of claim 1, the adjusted prediction information can be processed by the deep learning algorithm to derive improved prediction information; the adjusted prediction information is the adjusted corner point, and the improved prediction information is processed using the perspective transformation to obtain the improved corner point.

4. The device according to claim 1, wherein: The general prediction information is related to a plurality of corner points associated with the identity card, and The processor processes the general prediction information by perspective transformation-based processing to specifically determine information about the identity card.

5. The device according to claim 4, wherein: The identity card is associated with four corner points, and the general prediction information that can include p1, p2, p3 and p4 is related to the four corner points associated with the identity card, and A perspective transform based process is performed in conjunction with p1, p2, p3, and p4 to generate four defined corner points represented by p1', p2', p3', and p4'.

6. The device according to claim 5, wherein: The processor further processes the general prediction information p1 to p4 by adjustment-based processing to derive four adjusted corner points represented by p1(x,y), p2(x,y), p3(x,y) and p4(x,y), respectively, and Adjustment-based processing involves applying adjustment factors to p1 to p4.

7. The device according to claim 6, wherein: The ID card can be associated with a width w and a height h, and the ID card can be associated with a resolution defined based on "x" by "y", The adjustment factor is 0.1, and the adjusted corner point is: p1(x,y)=p1(x-0.1*w,y-0.1*h), p2(x,y)=p2(x+0.1*w,y-0.1*h), p3(x,y)=p3(x+0.1*w,y+0.1*h), p4(x,y)=p4(x-0.1*w,y+0.1*h).

8. The device according to claim 7, wherein: The processor transmits the adjusted corner points p1(x,y) to p4(x,y) to be processed by a deep learning algorithm to derive improved prediction information, and The processor further processes the refined prediction information by way of perspective transform based processing to derive four refined corner points denoted by pp1, pp2, pp3 and pp4 respectively.

9. The device according to claim 8, wherein: The targeted determination of information about the identification card is based on the plurality of defined corner points and / or the plurality of refined corner points.

10. A method for processing an identity card, comprising: A first stage of identity card detection, wherein general prediction information related to at least one corner point of the identity card is obtained based on a convolutional neural network; and The second stage is identity card detection, wherein information about the identity card is specifically determined based on the obtained general prediction information; Among them, based on the perspective transformation method, the general prediction information related to the at least one corner point is processed to obtain the corresponding defined corner point, and according to the adjustment factor, and the width and / or height associated with the identity card, the at least one defined corner point is adjusted respectively, and based on the perspective transformation method, the adjusted corner point is processed to obtain an improved corner point; the improved corner point is used to determine the information of the identity card.

11. The processing method according to claim 10, wherein: The convolutional neural network can be trained based on the first stage identity card detection and / or the second stage identity card detection.

Images