Identity data checking method, system and terminal

Abstract

The invention relates to an identity data verification method and system and a terminal, and relates to the field of identity recognition, and the method comprises the steps: obtaining a trigger signal of a preset call terminal; when the trigger signal is consistent with a preset call signal, obtaining call content of the call terminal; matching an identity keyword from a preset content database according to the call content; determining the identity of a call person according to the identity keyword; according to the identity of the call person, matching a reference person sound from a preset sound database, and obtaining a current person sound; and when the reference personnel sound is inconsistent with the current personnel sound, reporting an identity abnormity prompt. The method has the effect of improving the checking accuracy.

Description

Technical Field

[0001] The present invention relates to the field of identity recognition, and more particularly to a method, system and terminal for verifying identity data. Background Art

[0002] An effective identity data verification method refers to a set of systematic means for auditing and checking the accuracy, integrity, compliance and authenticity of identity data.

[0003] Currently, when a call terminal makes a call, it usually verifies the identity of the callers. During the verification, it generally analyzes the statistical data of the current callers in advance to obtain the identities of the current callers, and then takes corresponding actions based on the identities.

[0004] If the statistical data of the current callers is insufficient to support identity recognition, the call terminal will not perform other operations, thus reducing the accuracy of verification, which needs to be improved. Summary of the Invention

[0005] In order to improve the verification accuracy, the present invention provides a method, system and terminal for verifying identity data.

[0006] In a first aspect, the present invention provides a method for verifying identity data, adopting the following technical solution: A method for verifying identity data includes: Obtaining a trigger signal of a preset call terminal; When the trigger signal is consistent with a preset call signal, obtaining the call content of the call terminal; Matching identity keywords from a preset content database according to the call content; Determining the identity of the callers according to the identity keywords; Matching a reference person's voice from a preset voice database according to the identity of the callers, and obtaining the current person's voice; When the reference person's voice is inconsistent with the current person's voice, reporting an identity anomaly prompt.

[0007] By adopting the above technical solution, after receiving the trigger signal of the call terminal, if the signal is consistent with the call signal, the call content is obtained. Identity keywords are extracted from the content to clarify the identities of the callers, and then the reference person's voice is retrieved from the voice database based on these identities, while the current person's voice is collected. Once the two voices are inconsistent, an identity anomaly prompt is immediately reported, thereby improving the verification accuracy.

[0008] Optionally, it further includes an image recognition method: When the reference person's voice is consistent with the current person's voice, determining whether the call content contains preset sensitive words; When sensitive words are included in the call content, obtain the identification execution signal of the call terminal; When the identification execution signal is consistent with the preset activation identification signal, obtain the identification image information; When the identification image information does not include the preset accessory wearing feature, determine the current personnel feature according to the identification image information and the preset face feature; Determine the current personnel identity according to the current personnel feature; When the current personnel identity is inconsistent with the call personnel identity, report an identity anomaly prompt.

[0009] By adopting the above technical solution, after the voices of the reference personnel and the current personnel are consistent, the call content is further screened to determine whether it contains sensitive words. If sensitive words are included and the identification execution signal matches the activation identification signal, the identification image information is obtained. If there is no accessory wearing feature in the image, the current personnel feature is determined based on the face feature, and then their identity is clarified. Once it is found that there is a difference between the current personnel identity and the call personnel identity, an identity anomaly prompt is immediately reported. This effectively enhances the accuracy and reliability of identity verification and avoids potential risks.

[0010] Optionally, it further includes an identification method after wearing accessories: When the identification image information includes the preset accessory wearing feature, determine the accessory wearing position according to the identification image information, the accessory wearing feature, and the face feature; Determine the missing feature part and the remaining part feature according to the identification image information and the accessory wearing position; Match the detection angle value from the preset verification database according to the missing feature part; Control the call terminal to prompt the call personnel corresponding to the call personnel identity to adjust the angle according to the detection angle value; After the adjustment is completed, obtain the post-adjustment image information; When the post-adjustment image information includes the missing feature part, determine the missing part feature according to the post-adjustment image information, the missing feature part, and the detection angle value; Determine the complete face feature according to the missing part feature and the remaining part feature; Determine the current personnel identity according to the complete face feature.

[0011] By adopting the above technical solution, when identifying the jewelry wearing characteristics in the image information, first accurately locate the jewelry wearing position, thereby clarifying the missing feature parts caused by possible occlusion and the remaining part features. Then, according to the missing feature parts, match the appropriate detection angle value from the verification database, and guide the call personnel to adjust the angle through the call terminal. After the adjustment is completed, obtain the post-adjustment image information. If it contains the previously missing feature parts, then determine the missing part features by combining the post-adjustment image, the missing feature parts, and the detection angle value, and then integrate the remaining part features to form a complete face feature, and finally determine the identity of the current person, improving the recognition efficiency.

[0012] Optionally, the recognition method after wearing jewelry further includes: Based on the situation that the post-adjustment image information does not contain the missing feature parts, determine the jewelry occlusion range according to the recognition image information, the jewelry wearing characteristics, and the face features; Based on the situation that the jewelry occlusion range exceeds the preset influence occlusion range, determine the jewelry removal position according to the jewelry occlusion range, the influence occlusion range, and the jewelry wearing position; Control the call terminal to report a jewelry removal prompt according to the jewelry removal position; After the jewelry is removed, obtain the post-removal image information; Determine the removed part features according to the post-removal image information, the face features, and the jewelry removal position; Determine the identity of the current person according to the removed part features and the remaining part features.

[0013] By adopting the above technical solution, when the post-adjustment image information does not contain the missing feature parts, comprehensively judge the jewelry occlusion range according to the recognition image information, the jewelry wearing characteristics, and the face features. Once this range exceeds the preset influence occlusion range, it is possible to accurately locate the jewelry removal position by combining the jewelry occlusion range, the influence occlusion range, and the jewelry wearing position, and the call terminal reports a removal prompt to the call personnel. After the jewelry is removed, obtain the post-removal image information, determine the removed part features by virtue of this information, the face features, and the jewelry removal position, and then integrate the remaining part features to accurately judge the identity of the current person, further improving the verification accuracy.

[0014] Optionally, it further includes a face swap recognition method: Based on the situation that the identity of the current person is the same as that of the call personnel, control the playback device preset on the call terminal to play a swing detection prompt, and obtain the head swing image information; Determine whether the head swing image information contains the preset face swap features; Based on the situation that the head swing image information contains the face swap features, report an identity anomaly prompt; When there is no face-swapping feature in the head-shaking image information, match the facial color and neck color from the preset color database according to the head-shaking image information and the preset facial features; Match the color difference value from the preset difference database according to the facial color and neck color; When the color difference value exceeds the preset reference difference value, determine the color difference position according to the head-shaking image information, facial color and neck color, and obtain the difference image information of the color difference position; When the difference image information contains a face-swapping feature, report an identity anomaly prompt.

[0015] By adopting the above technical solution, after confirming that the identity of the current person is the same as that of the call participant, to further investigate risks, control the playback device on the call terminal to play a head-shaking detection prompt, and simultaneously obtain the head-shaking image information. Then, determine whether there is a face-swapping feature in the head-shaking image information. If it contains, directly report an identity anomaly prompt; if not, based on the head-shaking image information and facial features, match the facial color and neck color from the color database, and then obtain the color difference value from the difference database. When the color difference value exceeds the reference difference value, combine the head-shaking image information, etc. to determine the color difference position, obtain the difference image information at that position, and once it contains a face-swapping feature, also report an identity anomaly prompt, greatly improving the accuracy of face-swapping recognition.

[0016] Optionally, it also includes a light detection method: When the trigger signal is the same as the call signal, obtain the call image information of the call terminal; Determine whether the preset close feature is included in the call image information; When the call image information contains the close feature, determine the light arrival position according to the call image information and the close feature; Control the light emission device preset on the call terminal to emit light to the light arrival position, and obtain the light length value; When the light length value is the same as the preset reference length value, report an identity anomaly prompt.

[0017] By adopting the above technical solution, in the case where the trigger signal is the same as the call signal, obtain the call image information of the call terminal. When the close feature exists in the image information, determine the light arrival position by analyzing the image and the close feature, and then control the light emission device on the call terminal to emit light to this position to obtain the light length value. If this length value is the same as the preset reference length value, report an identity anomaly prompt, which can effectively improve the accuracy and reliability of identity verification.

[0018] Optionally, it also includes a method for obtaining the reference length value: Obtain the lens image information of the call terminal; When the lens image information contains a preset film application feature, determine the film thickness value according to the lens image information, the film application feature, and a preset reference object, and obtain the terminal model information of the call terminal; Determine the terminal screen parameters according to the terminal model information; Determine the light influence value according to the terminal screen parameters; Determine the screen thickness value according to the terminal screen parameters and the film thickness value; Determine the reference length value according to the light arrival position, the light influence value, the screen thickness value, and a preset light emission position.

[0019] By adopting the above technical solution, in order to accurately obtain the reference length value, first obtain the lens image information of the call terminal. When the film application feature appears in the image information, determine the film thickness value according to the lens image information, the film application feature, and the reference object, and at the same time obtain the terminal model information, thereby clarifying the terminal screen parameters. Then, calculate the light influence value and the screen thickness value by using the terminal screen parameters. Finally, comprehensively consider various factors such as the light arrival position, the light influence value, the screen thickness value, and the preset light emission position to determine the reference length value, ensuring the accuracy of the verification result.

[0020] Optionally, it further includes an algorithm formula for calculating the reference length value: L = L0 + (n * d * cosθ), where L is the reference length value, L0 is the light influence value, n is the refractive index of the film corresponding to the film application feature, d is the film thickness value, and θ is the refraction angle of the light.

[0021] By adopting the above technical solution, by using the above algorithm formula, the reference length value can be calculated more accurately, thereby improving the accuracy of the reference length value.

[0022] In a second aspect, the present application provides an identity data verification system, adopting the following technical solution: An identity data verification system includes: An acquisition module, configured to acquire a trigger signal, call content, current person's voice, recognition execution signal, recognition image information, adjusted image information, post-disconnection image information, head-turning image information, difference image information, call image information, light length value, lens image information, and terminal model information; A memory, configured to store the program of any one of the above identity data verification methods; A processor, configured to load and execute the program stored in the memory.

[0023] In a third aspect, the present application provides an intelligent terminal, adopting the following technical solution: An intelligent terminal includes a memory and a processor. A computer program capable of being loaded and executed by the processor to perform any of the above identity data verification methods is stored on the memory.

[0024] In summary, the present application includes at least one of the following beneficial technical effects: 1. After receiving a trigger signal from a call terminal, if the signal matches the call signal, the call content is obtained. Identity keywords are extracted from the content to clarify the identity of the callers. Then, based on this identity, the reference voice of the person is retrieved from the voice database, and the voice of the current person is collected simultaneously. Once the two voices are inconsistent, an identity anomaly prompt is immediately reported to improve the verification accuracy. 2. When there is a preset close - fitting feature in the image information, the position where the light arrives is determined by analyzing the image and the close - fitting feature. Then, the light - emitting device on the call terminal is controlled to emit light to this position, and the light length value is obtained. If the length value is the same as the reference length value, an identity anomaly prompt is reported, which can effectively improve the accuracy and reliability of identity verification. 3. When a film - sticking feature appears in the image information, the film - sticking thickness value is determined based on the lens image information, the film - sticking feature, and the reference object. At the same time, the terminal model information is obtained to clarify the terminal screen parameters. Then, the light influence value and the screen thickness value are calculated using the terminal screen parameters. Finally, considering multiple factors such as the light arrival position, the light influence value, the screen thickness value, and the preset light emission position, the reference length value is determined to ensure the accuracy of the verification result. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flowchart of a method for verifying identity data in an embodiment of the present invention; Figure 2 is a flowchart of an image recognition method in an embodiment of the present invention; Figure 3 is a flowchart of a recognition method after wearing accessories in an embodiment of the present invention Figure 1 ; Figure 4 is a flowchart of a recognition method after wearing accessories in an embodiment of the present invention Figure 2 ; Figure 5 is a flowchart of a face - swapping recognition method in an embodiment of the present invention; Figure 6 is a flowchart of a light detection method in an embodiment of the present invention; Figure 7 is a flowchart of a method for obtaining the reference length value in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] The present invention will be further described in detail below with reference to the drawings and embodiments.

[0027] An embodiment of this application discloses an identity data verification method.

[0028] Refer to Figure 1 , an identity data verification method includes the following steps: Step 100: Obtain a trigger signal of a preset call terminal.

[0029] A call terminal refers to a terminal used for online calls. In this embodiment, the call terminal is a mobile phone. A trigger signal refers to a signal triggered when a human operation is being performed on the call terminal. The trigger signal is obtained through a signal sensor on the call terminal.

[0030] Step 101: When the trigger signal is consistent with a preset call signal, obtain the call content of the call terminal.

[0031] A call signal refers to a signal when the call terminal is making a call. The call signal is set in advance by those skilled in the art and will not be elaborated here. The call content refers to the content during the call on the call terminal. The call content is obtained through a preset content recording APP in the call terminal. When the trigger signal is consistent with the call signal, it indicates that the call terminal is making a call, and the call content of the call terminal needs to be obtained for subsequent steps.

[0032] Step 102: Match identity keywords from a preset content database according to the call content.

[0033] Identity keywords refer to keywords used to assist in identifying the identity of the caller. The caller refers to the party initiating the call. Through the content database, corresponding identity keywords can be matched from the call content, which contains the corresponding relationship between the call content and the identity keywords. The content database is a manually set database and will not be elaborated here.

[0034] Step 103: Determine the identity of the call personnel according to the identity keywords.

[0035] The identity of the call personnel refers to the specific identity of the caller relative to the callee. Through a preset identity database, the identity of the call personnel corresponding to the identity keywords can be matched, which contains the corresponding relationship between the identity keywords and the identity of the call personnel. The identity database is a manually set database and will not be elaborated here.

[0036] Step 104: Match the reference personnel voice from a preset voice database according to the identity of the call personnel, and obtain the current personnel voice.

[0037] The reference person's voice refers to the voice characteristics corresponding to the identity of the call participant. The current person's voice refers to the voice characteristics of the current incoming caller. Through the voice database, the reference person's voice corresponding to the identity of the call participant can be matched. The voice database contains the correspondence between the identity of the call participant and the reference person's voice. The voice database is a manually set database and will not be elaborated here. The current person's voice is obtained through the voice recognition APP preset in the call terminal. After the reference person's voice is matched, the current person's voice needs to be obtained for subsequent steps.

[0038] Step 105: When the reference person's voice is inconsistent with the current person's voice, report an identity anomaly prompt.

[0039] The identity anomaly prompt refers to the prompt when the identity of the incoming caller does not match the identity of the call participant. The identity anomaly prompt is set in advance by those skilled in the art and will not be elaborated here. When the reference person's voice is inconsistent with the current person's voice, it indicates that the identity of the incoming caller does not match the identity of the call participant, and an identity anomaly prompt needs to be reported.

[0040] Refer to Figure 2 , the image recognition method includes the following steps: Step 200: When the reference person's voice is consistent with the current person's voice, determine whether the call content contains preset sensitive words.

[0041] Sensitive words refer to other sensitive words such as money involved in the call content. Sensitive words are set in advance by those skilled in the art and will not be elaborated here. When the reference person's voice is consistent with the current person's voice, it is necessary to further determine whether the call content contains sensitive words to know whether further identity recognition is required.

[0042] Step 201: When the call content contains sensitive words, obtain the recognition execution signal of the call terminal.

[0043] The recognition execution signal refers to the signal sent when the call terminal needs to perform further portrait recognition. The recognition execution signal is obtained through the signal transceiver on the call terminal. When the call content contains sensitive words, it indicates that further identity recognition is required, and the recognition execution signal of the call terminal needs to be obtained for subsequent steps.

[0044] Step 202: When the recognition execution signal is consistent with the preset recognition start signal, obtain the recognition image information.

[0045] The recognition start signal refers to the signal sent when the call terminal needs to perform further portrait recognition. The recognition start signal is set in advance by those skilled in the art and will not be elaborated here. The recognition image information refers to the image presented in front of the call terminal. The recognition image information is obtained by taking a photo with the camera preset on the call terminal.

[0046] When the recognition execution signal is consistent with the start recognition signal, it indicates that the call terminal needs to perform further portrait recognition and obtain recognition image information for subsequent steps.

[0047] Step 203: When the recognition image information does not contain the preset accessory wearing feature, determine the current person's feature according to the recognition image information and the preset face feature.

[0048] The accessory wearing feature refers to the feature when an accessory is worn on the face of the calling party. The face feature refers to various characteristics presented on the human face that are recognizable and can be used for distinguishing different individuals or for identity recognition, analysis, etc. Both the accessory wearing feature and the face feature are preset by those skilled in the art and will not be elaborated here. The current person's feature refers to the appearance feature on the face of the calling party. Through the preset feature database, the current person's feature corresponding to the recognition image information and the face feature can be matched, which contains the corresponding relationship between the recognition image information, the face feature, and the current person's feature. The feature database is a manually set database and will not be elaborated here.

[0049] When the recognition image information does not contain the accessory wearing feature, it indicates that no accessory is worn on the face of the calling party, and the current person's feature can be directly matched for subsequent steps.

[0050] Step 204: Determine the current person's identity according to the current person's feature.

[0051] The current person's identity refers to the identity information of the calling party. Through the identity database, the current person's identity corresponding to the current person's feature can be matched, which contains the corresponding relationship between the current person's feature and the current person's identity.

[0052] Step 205: When the current person's identity is inconsistent with the call person's identity, report an identity exception prompt.

[0053] When the current person's identity is inconsistent with the call person's identity, it indicates that the identity of the calling party does not match the call person's identity, and an identity exception prompt needs to be reported.

[0054] Refer to Figure 3 , the recognition method after wearing accessories includes the following steps: Step 300: When the recognition image information contains the preset accessory wearing feature, determine the accessory wearing position according to the recognition image information, the accessory wearing feature, and the face feature.

[0055] The position where the ornament is worn refers to the position where the ornament is worn on the face of the calling party. Through a preset ornament database, the ornament wearing position corresponding to the recognition image information, ornament wearing characteristics, and face characteristics can be matched. It contains the corresponding relationships among the recognition image information, ornament wearing characteristics, face characteristics, and ornament wearing position. The ornament database is a manually set database, which will not be elaborated here.

[0056] When the recognition image information contains ornament wearing characteristics, it indicates that the calling party wears an ornament on the face, and the ornament wearing position needs to be matched for subsequent steps.

[0057] Step 301: Determine the missing feature parts and the remaining part features based on the recognition image information and the ornament wearing position.

[0058] The missing feature parts refer to the feature parts that cannot be recognized on the face after being blocked by the ornament. The remaining part features refer to the feature parts that are not blocked by the ornament. Through the feature database, the missing feature parts and the remaining part features corresponding to the recognition image information and the ornament wearing position can be matched. It contains the corresponding relationships among the recognition image information, ornament wearing position, missing feature parts, and remaining part features.

[0059] Step 302: Match the detection angle value from the preset verification database according to the missing feature parts.

[0060] The detection angle value refers to the angle value formed by moving the call terminal so that there is a certain included angle between the call terminal and the missing feature parts. Through the verification database, the detection angle value corresponding to the missing feature parts can be matched. It contains the corresponding relationship between the missing feature parts and the detection angle value. The verification database is a manually set database, which will not be elaborated here.

[0061] Step 303: Control the call terminal to prompt the call personnel corresponding to the call personnel identity to adjust the angle according to the detection angle value.

[0062] Control the call terminal to prompt the call personnel corresponding to the call personnel identity to adjust the angle of the call terminal with the detection angle value for subsequent steps.

[0063] Step 304: After the adjustment is completed, obtain the post-adjustment image information.

[0064] The post-adjustment image information refers to the image presented in front of the call terminal after the call terminal is adjusted in angle. The post-adjustment image information is obtained by taking a photo with the camera on the call terminal. After the angle adjustment of the call terminal is completed, the post-adjustment image information needs to be obtained for subsequent steps.

[0065] Step 305: When the adjusted image information contains a missing feature part, determine the missing part feature based on the adjusted image information, the missing feature part, and the detection angle value.

[0066] The missing part feature refers to the appearance feature of the missing feature part. Through the feature database, the missing part feature corresponding to the adjusted image information, the missing feature part, and the detection angle value can be matched, which includes the corresponding relationship between the adjusted image information, the missing feature part, the detection angle value, and the missing part feature.

[0067] When the adjusted image information contains a missing feature part, it means that the missing part feature of the missing feature part can be further identified. Therefore, the missing part feature can be matched for subsequent steps.

[0068] Step 306: Determine the complete face feature based on the missing part feature and the remaining part feature.

[0069] The complete face feature refers to the complete face feature of the caller. The complete face feature can be obtained by combining the missing part feature and the remaining part feature.

[0070] Step 307: Determine the current person's identity based on the complete face feature.

[0071] The current person's identity refers to the identity information of the caller. Through the identity database, the current person's identity corresponding to the complete face feature can be matched, which includes the corresponding relationship between the complete face feature and the current person's identity.

[0072] Refer to Figure 4 , the recognition method after wearing ornaments further includes the following steps: Step 400: When the adjusted image information does not contain a missing feature part, determine the ornament occlusion range based on the recognition image information, the ornament wearing feature, and the face feature.

[0073] The ornament occlusion range refers to the range of the face of the caller blocked by the ornaments worn on the face. Through the preset range database, the ornament occlusion range corresponding to the recognition image information, the ornament wearing feature, and the face feature can be matched, which includes the corresponding relationship between the recognition image information, the ornament wearing feature, the face feature, and the ornament occlusion range. The range database is a database set by humans and will not be elaborated here.

[0074] When the adjusted image information does not contain a missing feature part, it means that even after angle adjustment, the missing part feature of the missing feature part cannot be recognized, and the ornament occlusion range needs to be matched for subsequent steps.

[0075] Step 401: When the occlusion range of the accessory exceeds the preset influence occlusion range, determine the position to remove the accessory based on the accessory occlusion range, the influence occlusion range, and the accessory wearing position.

[0076] The influence occlusion range refers to the minimum occlusion area where an influence starts to occur during the process of identifying the identity of the calling party. The influence occlusion range is set in advance by those skilled in the art and will not be elaborated here. The position to remove the accessory refers to the position on the face of the calling party where the accessory needs to be removed. Through the accessory database, the position to remove the accessory corresponding to the accessory occlusion range, the influence occlusion range, and the accessory wearing position can be matched, which includes the corresponding relationship between the accessory occlusion range, the influence occlusion range, the accessory wearing position, and the position to remove the accessory.

[0077] When the accessory occlusion range exceeds the influence occlusion range, it indicates that one or more accessories on the face of the calling party overly occlude the face, and some or all of the accessories need to be removed for identification. Therefore, the position to remove the accessory needs to be matched for subsequent steps.

[0078] Step 402: Control the call terminal to report a prompt to remove the accessory based on the position to remove the accessory.

[0079] The prompt to remove the accessory refers to a prompt to remove a certain or all accessories on the face. The prompt to remove the accessory is set in advance by those skilled in the art and will not be elaborated here. Control the call terminal to report the prompt to remove the accessory based on the position to remove the accessory.

[0080] Step 403: After the accessory is removed, obtain the post-removal image information.

[0081] The post-removal image information refers to the image of the calling party after the accessory is removed. The post-removal image information is obtained by taking a photo with a camera. After the accessory is removed, the post-removal image information needs to be obtained for subsequent steps.

[0082] Step 404: Determine the removed part features based on the post-removal image information, the facial features, and the position to remove the accessory.

[0083] The removed part features refer to the appearance features of the part after the accessory is removed. Through the feature database, the removed part features corresponding to the post-removal image information, the facial features, and the position to remove the accessory can be matched, which includes the corresponding relationship between the post-removal image information, the facial features, the position to remove the accessory, and the removed part features.

[0084] Step 405: Determine the current person's identity based on the removed part features and the remaining part features.

[0085] Through the identity database, the current person's identity corresponding to the removed part features and the remaining part features can be matched, which includes the corresponding relationship between the removed part features, the remaining part features, and the current person's identity.

[0086] Refer to Figure 5 , the face swapping recognition method includes the following steps: Step 500: When the current person's identity is the same as the identity of the call recipient, control the playback device preset on the call terminal to play a swing detection prompt, and obtain head swing image information.

[0087] The playback device refers to the device on the call terminal used for voice prompts. The swing detection prompt refers to a prompt used to detect the face swing of the caller. In this embodiment, face swing detection refers to the detection of the caller's head swinging up and down and left and right. The swing detection prompt is set in advance by those skilled in the art and will not be elaborated here. The head swing image information refers to the image when the caller performs face swing detection. The head swing image information is obtained by taking pictures with a camera.

[0088] When the current person's identity is the same as the identity of the call recipient, to avoid a face mask, it is necessary to control the playback device on the call terminal to play a swing detection prompt and obtain head swing image information for subsequent steps.

[0089] Step 501: Determine whether the head swing image information contains a preset face swapping feature.

[0090] The face swapping feature refers to the feature when the caller wears a face swapping mask on the face. The face swapping feature is set in advance by those skilled in the art and will not be elaborated here. By judging whether the head swing image information contains the face swapping feature, it can be known whether the identity of the caller is fake.

[0091] Step 502: When the head swing image information contains the face swapping feature, report an identity anomaly prompt.

[0092] When the head swing image information contains the face swapping feature, it means that the identity of the caller is fake, and an identity anomaly prompt needs to be reported.

[0093] Step 503: When the head swing image information does not contain the face swapping feature, match the facial color and neck color from the preset color database according to the head swing image information and the preset face features.

[0094] Face features refer to various characteristics presented by a person's face that are recognizable and can be used for distinguishing different individuals or for identity recognition, analysis, etc. Face features are set in advance by those skilled in the art and will not be elaborated here. The facial color refers to the color of the caller's face. The neck color refers to the color of the caller's neck. Through the color database, the facial color and neck color corresponding to the head swing image information and the face features can be matched, which includes the corresponding relationship between the head swing image information, face features, facial color, and neck color. The color database is a manually set database and will not be elaborated here.

[0095] When the head-turning image information does not contain face-swapping features, in order to further detect the identity of the caller, it is necessary to match the facial color and neck color for subsequent steps.

[0096] Step 504: Match the color difference value from the preset difference database according to the facial color and neck color.

[0097] The color difference value refers to the quantitative value of the degree of difference between the facial color and the neck color. Through the difference database, the color difference value corresponding to the facial color and the neck color can be matched, which contains the corresponding relationship between the facial color, the neck color, and the color difference value. The difference database is a manually set database and will not be elaborated here.

[0098] Step 505: When the color difference value exceeds the preset reference difference value, determine the color difference position based on the head-turning image information, facial color, and neck color, and obtain the difference image information of the color difference position.

[0099] The reference difference value refers to the maximum value that the color difference value is allowed to reach. The reference difference value is set in advance by those skilled in the art and will not be elaborated here. The color difference position refers to the position where the difference between the facial color and the neck color appears. The facial and neck colors are distinguished through the head-turning image information, and then the specific position where the colors of the two are different is determined and defined as the color difference position. The difference image information refers to the image of the color difference position. The difference image information is obtained by taking a photo with a camera.

[0100] When the color difference value exceeds the reference difference value, it indicates that the color difference between the face and the neck is too large. It is necessary to first determine the color difference position and then obtain the difference image information for subsequent steps.

[0101] Step 506: Report an identity anomaly prompt when the difference image information contains face-swapping features.

[0102] When the difference image information contains face-swapping features, it indicates that the identity of the caller is fake, and an identity anomaly prompt needs to be reported.

[0103] Refer to Figure 6 , the light detection method includes the following steps: Step 600: Obtain the call image information of the call terminal when the trigger signal is consistent with the call signal.

[0104] The call image information refers to the image presented in front of the call terminal during a call. The call image information is obtained by taking a photo with a camera. When the trigger signal is consistent with the call signal, it indicates that the call terminal is in a call, and it is necessary to obtain the call image information for subsequent steps.

[0105] Step 601: Determine whether the preset close contact feature is included in the call image information.

[0106] The close contact feature refers to the feature that the call terminal is in close contact with the face. The close contact feature is preset by those skilled in the art and will not be elaborated here. By determining whether the close contact feature is included in the call image information, it can be known whether the call terminal is in close contact with the human face.

[0107] Step 602: When the close contact feature is included in the call image information, determine the light arrival position according to the call image information and the close contact feature.

[0108] The light arrival position refers to the position where the light used to detect whether a face mask is worn on the human face is emitted. Through the preset light database, the light arrival position corresponding to the call image information and the close contact feature can be matched, which includes the corresponding relationship between the call image information, the close contact feature, and the light arrival position. The light database is a manually set database and will not be elaborated here.

[0109] When the close contact feature is included in the call image information, it indicates that the call terminal is in close contact with the human face, and the light arrival position needs to be matched for subsequent steps.

[0110] Step 603: Control the light emitting device preset on the call terminal to emit light to the light arrival position and obtain the light length value.

[0111] The light emitting device refers to the device used to emit the light for detecting whether a face mask is worn on the human face. The light length value refers to the distance that the light travels after being emitted and reaches the human face. The light length value is measured and obtained by the light sensor preset on the call terminal. Control the light emitting device on the call terminal to emit light to the light arrival position and obtain the light length value for subsequent steps.

[0112] Step 604: When the light length value is consistent with the preset reference length value, report an identity anomaly prompt.

[0113] The reference length value refers to the length value that the light should have when there are no other interfering factors. The method for obtaining the reference length value will be described in detail in subsequent steps 700 to 705 and will not be elaborated here. Since oil will be produced on the human face, an oil film will be formed, which will cause the light length value to be inconsistent with the reference length value. If the light length value is consistent with the reference length value, it indicates that there is no oil on the currently recognized human face, and further indicates that the currently recognized human face is a face mask, and an identity anomaly prompt needs to be reported.

[0114] Refer to Figure 7 , the method for obtaining the reference length value includes the following steps: Step 700: Obtain the lens image information of the call terminal.

[0115] The lens image information refers to the image around the lens of the call terminal. The lens image information is obtained by taking a photo with the camera.

[0116] Step 701: When the lens image information contains a preset film - sticking feature, determine the film - sticking thickness value based on the lens image information, the film - sticking feature, and a preset reference object, and obtain the terminal model information of the call terminal.

[0117] The film - sticking feature refers to the feature when a protective film is stuck on the call terminal. The film - sticking feature is set in advance by those skilled in the art and will not be elaborated here. The reference object is an object used to assist in measuring the thickness value of the protective film. The size and position of the reference object are set in advance by those skilled in the art and will not be elaborated here, and the lens image information contains the reference object. The film - sticking thickness value refers to the thickness value of the protective film. The terminal model information refers to the specific model of the call terminal. The terminal model information is obtained by querying the model introduction APP in the call terminal.

[0118] When the lens image information contains the film - sticking feature, it indicates that a protective film is stuck on the screen of the call terminal. It is necessary to first determine the film - sticking thickness value and then obtain the terminal model information of the call terminal for subsequent steps.

[0119] Step 702: Determine the terminal screen parameters based on the terminal model information.

[0120] The terminal screen parameters refer to a series of indicators used to describe the screen characteristics of the call terminal, which include size, resolution, pixel density, screen ratio, screen thickness, refresh rate, color parameters, brightness, and contrast. Through a preset terminal database, the terminal screen parameters corresponding to the terminal model information can be matched. The terminal database contains the corresponding relationship between the terminal model information and the terminal screen parameters. The terminal database is a manually set database and will not be elaborated here.

[0121] Step 703: Determine the light influence value based on the terminal screen parameters.

[0122] The light influence value refers to the influence value of different screen parameters on the light emitted by the light - emitting device. Through the light database, the light influence value corresponding to the terminal screen parameters can be matched, which contains the corresponding relationship between the terminal screen parameters and the light influence value.

[0123] Step 704: Determine the screen thickness value based on the terminal screen parameters and the film - sticking thickness value.

[0124] The screen thickness value refers to the sum of the screen thickness of the call terminal itself and the film - sticking thickness value of the stuck film. The screen thickness value can be obtained by adding the screen thickness of the call terminal itself and the film - sticking thickness value.

[0125] Step 705: Determine a reference length value based on the light arrival position, the light influence value, the screen thickness value, and a preset light emission position.

[0126] The light emission position refers to the position of the light emitted by the light emission device. The light emission position is preset by those skilled in the art and will not be elaborated here. The reference length value can be calculated through the algorithm formula L = L0 + (n * d * cosθ), where L is the reference length value, L0 is the light influence value, n is the refractive index of the film corresponding to the film characteristics, d is the film thickness value, and θ is the light refraction angle. n and θ are measured in advance by those skilled in the art and will not be elaborated here.

[0127] Based on the same inventive concept, an embodiment of the present invention provides an identity data verification system, including: An acquisition module, configured to acquire a trigger signal, a call content, the current person's voice, an identification execution signal, identification image information, a call number, call record information, the current IP location, a head shaking image information, difference image information, call image information, a light length value, lens image information, and terminal model information; A memory, configured to store a program of an identity data verification method; A processor, configured to load and execute the program stored in the memory.

[0128] Based on the same inventive concept, an embodiment of the present invention provides an intelligent terminal, including a memory and a processor, and a computer program capable of being loaded and executed by the processor is stored on the memory for an identity data verification method.

[0129] Those skilled in the art can clearly understand that for the convenience and brevity of description, only the above division of each functional module is used as an example for illustration. In practical applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here.

[0130] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the inventive concept of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. A method for verifying identity data, characterized in that: include: Obtaining a trigger signal from a preset call terminal; When the trigger signal is consistent with the preset call signal, the call content of the call terminal is obtained; Match identity keywords from a preset content database based on the call content; Determine the identity of the caller based on identity keywords; According to the identity of the caller, a reference person's voice is matched from a preset voice database, and the current person's voice is obtained; When the voice of the benchmark person is inconsistent with the voice of the current person, an identity anomaly prompt is reported.

2. The identity data verification method according to claim 1, characterized in that: Also includes image recognition methods: When the voice of the reference person is consistent with the voice of the current person, determine whether the call content contains preset sensitive words; When the call content contains sensitive words, obtain the recognition execution signal of the call terminal; When the recognition execution signal is consistent with the preset start recognition signal, the recognition image information is obtained; When the recognition image information does not contain the preset jewelry wearing features, the current person features are determined based on the recognition image information and the preset facial features; Determine the identity of the current person based on the current person's characteristics; When the current person's identity is inconsistent with the calling person's identity, an identity anomaly prompt is reported.

3. The identity data verification method according to claim 2, characterized in that: It also includes identification methods after wearing jewelry: When the recognition image information contains preset jewelry wearing features, the jewelry wearing position is determined according to the recognition image information, the jewelry wearing features and the facial features; Determine the missing feature parts and the remaining feature parts according to the recognition image information and the wearing position of the jewelry; Match the detection angle value from the preset verification database according to the missing feature parts; According to the detected angle value, the call terminal is controlled to prompt the caller corresponding to the caller's identity to adjust the angle; When the adjustment is completed, the adjusted image information is obtained; When the adjusted image information contains a missing feature part, the missing feature part is determined according to the adjusted image information, the missing feature part and the detection angle value; Determine complete facial features based on missing and remaining features; Determine the current person's identity based on complete facial features.

4. The identity data verification method according to claim 3, characterized in that: Other identification methods after wearing jewelry include: When the adjusted image information does not contain the missing feature part, the jewelry occlusion range is determined according to the recognition image information, jewelry wearing features and facial features; When the jewelry occlusion range exceeds the preset influence occlusion range, the jewelry removal position is determined according to the jewelry occlusion range, the influence occlusion range and the jewelry wearing position; Control the call terminal to report the jewelry removal prompt according to the jewelry removal position; When the jewelry is removed, the image information after removal is obtained; Determine the features of the removed part based on the image information after removal, the facial features and the location of the removed jewelry; The identity of the current person is determined based on the characteristics of the removed parts and the remaining parts.

5. The identity data verification method according to claim 2, characterized in that: Also includes face recognition method: When the current person's identity is consistent with the calling person's identity, the playback device preset on the calling terminal is controlled to play the swing detection prompt and obtain the head swing image information; Determine whether the head shaking image information contains a preset face-changing feature; When the head shaking image information contains face-changing features, an identity anomaly prompt is reported; When the head shaking image information does not contain face-changing features, the facial color and the neck color are matched from a preset color database according to the head shaking image information and preset facial features; Match the color difference value from the preset difference database according to the face color and the neck color; When the color difference value exceeds a preset reference difference value, the color difference position is determined according to the head shaking image information, the facial color and the neck color, and the difference image information of the color difference position is obtained; When the difference image information contains face-changing features, an identity anomaly prompt is reported.

6. The identity data verification method according to claim 1, characterized in that: Also includes light detection methods: When the trigger signal is consistent with the call signal, the call image information of the call terminal is obtained; Determining whether the call image information contains a preset close contact feature; When the call image information includes a close contact feature, determining the light arrival position according to the call image information and the close contact feature; Control the light emitting device preset on the call terminal to emit light to the light arrival position, and obtain the light length value; When the light length value is consistent with the preset reference length value, an identity abnormality prompt is reported.

7. The identity data verification method according to claim 6, characterized in that: It also includes a method for obtaining the reference length value: Obtain the lens image information of the call terminal; When the lens image information contains preset film features, determine the film thickness value according to the lens image information, the film features and the preset reference object, and obtain the terminal model information of the call terminal; Determine terminal screen parameters according to terminal model information; Determine the light impact value based on the terminal screen parameters; Determine the screen thickness value according to the terminal screen parameters and the film thickness value; The reference length value is determined according to the light arrival position, the light impact value, the screen thickness value and the preset light emission position.

8. The identity data verification method according to claim 7, characterized in that: Also included is the algorithm formula for calculating the base length value: L=L0+(n*d*cosθ), where L is the reference length value, L0 is the light impact value, n is the refractive index of the film corresponding to the film characteristics, d is the film thickness value, and θ is the refraction angle of the light.

9. An identity data verification system, characterized in that: include: The acquisition module is used to obtain the trigger signal, call content, current person's voice, recognition execution signal, recognition image information, adjusted image information, removed image information, head shaking image information, difference image information, call image information, light length value, lens image information and terminal model information; A memory for storing a program of an identity data verification method according to any one of claims 1 to 8; The processor is used to load, execute and implement the program stored in the memory.

10. An intelligent terminal, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executes an identity data verification method as claimed in any one of claims 1 to 8.

Images