Iris template entry and authentication method, device and medium for terminal multi-modal multiplexing

Abstract

This application relates to a method for iris template input and authentication for multi-morphological reuse in eye-tracking terminals. By introducing a mapping relationship between terminal morphology encoding and gaze direction, the adaptability and stability of iris image acquisition are significantly improved. This method ensures stable acquisition of iris images under different terminal morphology conditions, effectively solving authentication failures or incorrect matching problems caused by device differences. During the input and authentication process, the system ensures high accuracy of authentication by accurately matching the terminal morphology encoding and gaze direction, avoiding the computational burden and latency caused by full comparison. Through the "one-time input, multi-terminal reuse" scheme, this application reduces repetitive input, lowers maintenance costs, and improves user experience. Users do not need to input templates separately for each terminal, thus achieving a more efficient and flexible iris authentication process.

Description

Technical Field

[0001] This application relates to the field of eye-tracking technology, and in particular to a method, device and medium for iris template input and authentication for multi-form reuse of eye-tracking terminals. Background Technology

[0002] Iris recognition, due to its stable texture features and high discriminative power, is widely used in access control gates, attendance terminals, and identity verification devices. Existing iris recognition systems typically include two stages: registration and authentication. In the registration stage, the user is guided to complete iris image acquisition in a predetermined posture, generating an iris template. In the authentication stage, the iris image to be authenticated is acquired and matched against the stored iris template to output the authentication result. With the development of human-computer interaction technology, some terminals have incorporated eye-tracking capabilities, estimating the user's gaze direction and fixation point to assist in guiding the acquisition process, thereby improving acquisition efficiency and usability.

[0003] However, in eye-tracking terminals, due to differences in industrial design, screen size, module layout, and installation structure, the installation position of the iris acquisition and imaging module relative to the gaze interface varies. For example, the imaging module may be located above or below the gaze interface, or even without a gaze interface at all. These different configurations result in varying off-axis imaging gaze directions for users under the condition of "normal gaze at the gaze interface," leading to differences in the imaging angle, visible texture area, and occlusion of the iris image. This, in turn, affects template stability and matching accuracy. For the same user, when switching between terminals of different configurations, using a single template or a fixed guidance strategy for data entry can easily lead to decreased authentication success rates and increased false rejections.

[0004] To accommodate off-axis differences, common approaches in existing technologies include: repeatedly inputting data for different terminals; storing multiple templates during the input phase and performing a full comparison during the authentication phase; or relying on eye tracking / pose estimation results to select a template for matching during the authentication phase. These solutions still have shortcomings in practical deployment: repeated input increases maintenance costs and user burden; full comparison leads to increased computational load and response latency, and the increased number of templates can easily introduce the risk of mismatches; while relying solely on real-time gaze estimation for template selection is susceptible to factors such as instantaneous eye movements, blinking occlusion, estimation errors, and inconsistencies in the geometric definitions of different terminals, leading to unstable or incorrect template selection. Especially in application scenarios where there are multiple module installation forms within the same product family and the goal is "one-time input, multiple terminal reuse," there is still a lack of a technical solution that can incorporate terminal form differences into the input and authentication loop in a clear and transferable manner, while ensuring selection accuracy and reducing comparison overhead. Summary of the Invention

[0005] The purpose of this application is to provide a technical solution that can incorporate terminal form differences into the input and authentication closed loop in a clear and transferable manner, while reducing comparison overhead while ensuring selection accuracy.

[0006] According to one aspect of this application, a method for iris template input and authentication for multi-form reuse in eye-tracking terminals is provided, comprising the following steps: S10. Provide an input device, the input device including an eye-tracking unit and an iris acquisition and imaging unit, and the input device having a gaze interface for presenting guide points; S20. Establish terminal form code and establish a preset mapping relationship between the terminal form code and the target gaze direction. The terminal form code represents the vertical position relationship between the iris acquisition imaging module of the terminal to be authenticated and the gaze interface of the terminal to be authenticated. S30. During the input phase, input operations are performed sequentially for at least two different terminal form codes: the target gaze direction corresponding to the preset mapping relationship is determined based on the current terminal form code, and a guide point is presented on the gaze interface of the input device; the gaze direction parameter is output by the eye-tracking unit and compared with the target gaze direction; when the allowable deviation is met, the iris acquisition and imaging unit is driven to acquire an iris image and generate an iris template; the iris template is then associated with the current terminal form code and stored. S40. During the authentication phase, the terminal morphology code is obtained from the terminal to be authenticated; the eye-tracking unit outputs the gaze direction parameters at the time of authentication and compares them with the target gaze direction corresponding to the preset mapping relationship. When the allowable deviation is met, the iris template corresponding to the obtained terminal morphology code is selected from the stored iris templates and matched with the iris template generated from the iris image to be authenticated to output the authentication result.

[0007] More preferably, the terminal to be authenticated includes a gaze interface, an iris acquisition and imaging module, and a control unit electrically connected to the iris acquisition and imaging module, wherein the control unit includes a non-volatile memory and a communication interface; The terminal form code is written into the non-volatile memory, and the terminal form code includes at least an upper form code and a lower form code. The upper form code and the lower form code correspond to the upper and lower installation forms of the iris acquisition imaging module relative to the gaze interface of the terminal to be authenticated, respectively. During the authentication stage, the terminal form code is output through the communication interface.

[0008] More preferably, the input device includes a mapping table storage area, which stores mapping table entries. Each mapping table entry includes a terminal morphology encoding field and a target gaze direction field, and the terminal morphology encoding field corresponds one-to-one with the target gaze direction field. During the input phase and the authentication phase, the target gaze direction field is retrieved based on the terminal form encoding field to determine the target gaze direction.

[0009] More preferably, the gaze direction parameter output by the eye-tracking unit includes a gaze direction vector, and the target gaze direction is represented by the target gaze direction vector; The input device calculates the angle between the gaze direction vector and the target gaze direction vector, and drives the iris acquisition imaging unit to acquire the iris image when the angle is less than a preset angle threshold and is continuously satisfied within a preset time window. Within the preset time window, when the blink state parameters output by the eye-tracking unit indicate that a blink or eye occlusion has occurred, the driving of the iris acquisition imaging unit is stopped and the timing of the preset time window is restarted.

[0010] More preferably, the data entry device includes a quality assessment unit; During the input phase, for the same terminal form encoding, the iris acquisition and imaging unit acquires multiple iris images at different time points; The quality assessment unit obtains quality indicators for each iris image and compares them with a preset quality threshold. When the quality indicators meet the preset quality threshold, the iris template is generated based on the corresponding iris image, and the iris template is associated with the terminal morphology code and stored. If the quality indicators do not meet the preset quality threshold, the presentation of the guide point is maintained and the iris image continues to be acquired.

[0011] More preferably, the input device or the terminal to be authenticated includes a storage structure, the storage structure including a template storage area and an index storage area; The template storage area stores the iris template; The index storage area stores index entries, each index entry including a terminal morphology encoding field and a template identifier field, with the terminal morphology encoding field and the template identifier field having a one-to-many correspondence; During the authentication phase, the target iris template set corresponding to the template identifier field is determined in the index storage area based on the terminal morphology encoding field, and the matching is performed only on the target iris template set to output the authentication result.

[0012] More preferably, during the authentication stage, when the terminal form code output by the terminal to be authenticated is missing or does not meet the preset validity conditions, the input device determines the candidate terminal form code based on the deviation between the gaze direction parameter and each target gaze direction in the preset mapping relationship, and selects the iris template to perform the matching according to the candidate terminal form code.

[0013] More preferably, the storage structure includes an update rule storage area, which stores update rules corresponding to the terminal form code; During the input phase, when a new iris template is generated for the same terminal morphology code, the new iris template is written into the template storage area according to the update rule, and the template identifier field of the index entry corresponding to the terminal morphology code field is updated synchronously.

[0014] In one specific embodiment of this application, a computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to perform the steps of any of the methods described herein.

[0015] In one specific embodiment of this application, a computer device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of any of the methods described.

[0016] This application has the following beneficial effects: This application effectively improves the adaptability and stability of iris image acquisition by introducing a mapping relationship between terminal morphology encoding and gaze direction. This application ensures the stability of iris image acquisition under different terminal morphology configurations, resolving authentication failures or incorrect matching issues caused by device differences. During the input and authentication process, the system ensures high accuracy of authentication by accurately matching terminal morphology encoding and gaze direction, while avoiding the computational burden and latency caused by full comparison. Furthermore, this application reduces maintenance costs and improves user experience through a "one-time input, multi-terminal reuse" technical solution. The system design eliminates the need for users to input templates separately for each terminal, thus achieving a more efficient and flexible iris authentication process. This application provides an efficient, accurate, and highly adaptable iris template input and authentication method that significantly improves system efficiency and reduces maintenance costs while ensuring authentication accuracy. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is described in one embodiment of this application. Detailed Implementation

[0019] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings. Preferred embodiments of this application are shown in the drawings. However, this application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0020] Please refer to Figure 1 This embodiment provides a method for iris template input and authentication for multi-form reuse in eye-tracking terminals, which is described in the following steps.

[0021] S10. Provide an input device. The input device includes an eye-tracking unit and an iris acquisition and imaging unit, and the input device has a gaze interface for presenting guide points. In one embodiment, the input device further includes a processor and a memory. The processor is used to perform processing such as determining the target gaze direction corresponding to the terminal morphology code, judging the allowable deviation, triggering iris acquisition, generating an iris template, and matching the template. The memory is used to store preset mapping relationships and iris templates corresponding to the terminal morphology code.

[0022] S20. Establish a terminal form code and a preset mapping relationship between the terminal form code and the target gaze direction. The terminal form code represents the vertical position of the iris acquisition imaging module of the terminal to be authenticated relative to the gaze interface of the terminal to be authenticated. For ease of implementation, in one embodiment, the preset mapping relationship is stored in the memory of the input device in the form of a mapping table. The mapping table includes multiple mapping table entries, each including at least a terminal form code field and a target gaze direction field, and the terminal form code field and the target gaze direction field correspond one-to-one; wherein, the target gaze direction field can be represented as "target gaze point position" or "target gaze direction parameter", used to represent the target gaze direction that the operator should achieve under the corresponding terminal form. The input device can retrieve the corresponding target gaze direction field based on the terminal form code field in both the input stage and the authentication stage, thereby ensuring the consistency of the terminal form code, the preset mapping relationship, and the target gaze direction at different stages.

[0023] S30. In the input stage, input operations are sequentially performed for at least two different terminal form codes. Taking the current terminal form code as c, the input device determines the corresponding target line of sight direction in the preset mapping relationship according to the current terminal form code c, and presents a guiding point on the gaze interface of the input device to guide the operator's line of sight to approach the target line of sight direction; subsequently, the line of sight direction parameter is output by the eye movement tracking unit, and the line of sight direction parameter is compared with the target line of sight direction. When the allowable deviation is satisfied, the iris acquisition and imaging unit is driven to acquire an iris image and generate an iris template, and the iris template is established corresponding to the current terminal form code c and stored.

[0024] In one embodiment, for the sake of easy understanding and implementation, the "comparison between the line of sight direction parameter and the target line of sight direction" can be specifically embodied as "calculating the deviation angle between the two and comparing it with a threshold value". For example, define: angle_now: the included angle between the current line of sight direction and the target line of sight direction; angle_limit: the allowable deviation threshold (preset angle threshold).

[0025] When angle_now < angle_limit is satisfied, it is determined that the allowable deviation is satisfied. Among them, angle_now can be calculated from the line of sight direction parameter output by the eye movement tracking unit and the target line of sight direction field, and the specific calculation method can be implemented by using the conventional included angle calculation method in the art.

[0026] To improve the stability of the acquisition trigger, in one embodiment, a time window time_window is set, and it is required that the allowable deviation be continuously satisfied within the time window to trigger the acquisition. Taking the output frequency of the eye movement tracking unit as fps, the required continuous satisfaction times count_need can be determined as follows:

[0027] Among them, ceil(.) is the ceiling function. The trigger rule can be: when it is continuously determined that angle_now < angle_limit for count_need times, the iris acquisition and imaging unit is driven to acquire an iris image; if angle_now < angle_limit is not satisfied某次 during the process, the continuous count is cleared and the counting starts again. Through the above method, it is possible to avoid false triggering of the acquisition caused by short-term jitter.

[0028] After triggering the acquisition, the iris acquisition imaging unit acquires an iris image, and the recording device performs template generation processing on the iris image to generate an iris template. The template generation processing may include steps such as iris region localization, normalization, and feature encoding to obtain an iris template in the form of a fixed-length feature vector or bit string. After performing the above recording operation for at least two different terminal morphology codes, the recording device can generate iris template data corresponding to each terminal morphology code, thereby providing a basis for selecting a template according to the terminal morphology code during the authentication stage.

[0029] S40. During the authentication phase, the terminal morphology code is obtained from the terminal to be authenticated; the eye-tracking unit outputs the gaze direction parameter at the moment of authentication and compares it with the target gaze direction corresponding to the preset mapping relationship. When the allowable deviation is met, the iris template corresponding to the obtained terminal morphology code is selected from the stored iris templates and matched with the iris template generated from the iris image to be authenticated to output the authentication result.

[0030] In one embodiment, the iris image to be authenticated generates an iris template `template_q`. The input device, based on the acquired terminal morphology code, selects a reference iris template or a set of reference iris templates `TemplateSet(c)` corresponding to that terminal morphology code from storage, and calculates the difference value between the iris template to be authenticated and the reference iris template; the smaller the difference value, the higher the similarity. Let: diff(a,b): The difference between the two iris templates; diff_limit: Matching threshold.

[0031] The authentication result can then be output using the following judgment rules: min_{template_rinTemplateSet(c)}diff(template_q,template_r)<diff_limit-> Certification passed Otherwise -> authentication failed.

[0032] The `diff(.)` function can be implemented using Hamming distance, Euclidean distance, or difference values ​​calculated based on similarity, which are all equivalent implementation methods available in this field. By selecting the corresponding iris template for matching based on the terminal morphology encoding during the authentication phase, and triggering acquisition and matching when the gaze direction meets the allowable deviation, the stability of iris acquisition and authentication can be improved under the condition of multi-morphology reuse of terminals, and the risk of mismatch caused by template mixing between different terminal morphologies can be reduced.

[0033] Based on the above embodiments, to implement the process of "obtaining terminal morphology code from the terminal to be authenticated" in the authentication stage S40, in one embodiment, the terminal to be authenticated includes a gaze interface, an iris acquisition and imaging module, and a control unit electrically connected to the iris acquisition and imaging module. The control unit is used to power and control the iris acquisition and imaging module, and to communicate with the input device. The control unit includes a non-volatile memory and a communication interface, wherein the non-volatile memory is used to store the terminal morphology code, and the communication interface is used to output the terminal morphology code to the input device during the authentication stage, enabling the input device to perform subsequent processing based on the acquired terminal morphology code.

[0034] In one embodiment, the terminal form code is used to characterize the vertical position of the iris acquisition and imaging module relative to the gaze interface of the terminal to be authenticated, and the terminal form code includes at least an upper form code and a lower form code. The upper form code corresponds to the upper installation position of the iris acquisition and imaging module relative to the gaze interface, and the lower form code corresponds to the lower installation position of the iris acquisition and imaging module relative to the gaze interface. For ease of implementation, the terminal form code can be represented by discrete values. For example, the terminal form code can be denoted as code, and it can be specified that the upper form code is code=1 and the lower form code is code=2. The above values ​​are only examples, and bytecode, enumeration values, or bit fields can also be used to represent it. The upper and lower positions only need to be distinguishable.

[0035] In an optional implementation, the terminal morphology code can be automatically determined and written to the non-volatile memory based on installation morphology parameters. For example, the vertical coordinate of the center of the gaze interface can be taken as y_disp in the terminal coordinate system, and the vertical coordinate of the center of the iris acquisition and imaging module can be taken as y_cam, and the vertical relative position difference between the two can be calculated: delta=y_cam-y_disp When delta > 0, it is determined to be an upper installation configuration and code = 1; when delta <= 0, it is determined to be a lower installation configuration and code = 2. The y_disp and y_cam values ​​can be derived from factory assembly parameters, installation tooling measurements, or the terminal's internal configuration table; the specific acquisition method is not limited. This method converts the "vertical positional relationship of the iris acquisition imaging module relative to the gaze interface" into a terminal configuration code and saves it.

[0036] In one embodiment, the terminal configuration code is written into the non-volatile memory to achieve power-off retention. To improve reproducibility and reliability, both the encoding field and the check field can be stored in the non-volatile memory simultaneously. For example, a storage address `addr_code` is set to store the code, and a storage address `addr_chk` is set to store the check value `chk`, and written as follows: NVM[addr_code]=code chk=(codeXOR0xFF) NVM[addr_chk]=chk Here, XOR represents the exclusive OR operation; the above verification method is only an example, and equivalent verification can also be achieved using CRC8, CRC16, etc. Correspondingly, during the authentication phase, the data can first be read from and verified using non-volatile memory: code_read = NVM[addr_code] chk_read=NVM[addr_chk] (valid)=(chk_read==(code_readXOR0xFF))AND(code_readin{1,2}) When valid is true, the terminal form code is determined to be valid and output is allowed; when valid is false, the terminal form code is determined to be missing or invalid, thereby avoiding incorrect output encoding due to storage abnormalities.

[0037] In the authentication phase S40, the terminal to be authenticated outputs the terminal form code through the communication interface. The communication interface can be a wired or wireless communication interface, specifically USB, serial port, wired network, Bluetooth, or near-field communication, etc., with no limitation on the communication method. The output method can be one of two equivalent mechanisms: active reporting or request-response. For example, in the request-response mode, after the input device sends an encoding read request, the terminal to be authenticated reads code_read from the non-volatile memory and returns data containing the terminal form code field through the communication interface; to facilitate correct parsing by the input device, the output data can be organized into a data frame containing the "terminal identifier field dev_id" and the "form code field code", and optionally carry a verification field, for example: Frame=[dev_id,code,(dev_id+code)mod256] Here, "(dev_id+code)mod256" is an example verification field. After receiving the data, the input device can calculate and verify consistency according to the same rule. If the verification passes, the code is accepted as the terminal form code. Thus, the input device can complete the action of "obtaining the terminal form code from the terminal to be authenticated" in S40, thereby ensuring a consistent correspondence and reproducibility between the writing, storage, and output of the terminal form code on the terminal to be authenticated side, and the receiving and use on the input device side.

[0038] In one embodiment, the input device includes a mapping table storage area for storing pre-configured mapping table entries. Each mapping table entry includes at least a terminal form encoding field and a target gaze direction field, with a one-to-one correspondence between the terminal form encoding field and the target gaze direction field. The terminal form encoding field indicates the form category of the terminal to be authenticated (e.g., to distinguish between upper and lower installation forms of the iris acquisition imaging module relative to the gaze interface), and the target gaze direction field indicates the desired target gaze direction under the corresponding terminal form, thereby enabling the input device to call different target gaze directions for guidance and comparison under different terminal forms.

[0039] For ease of implementation, in one embodiment, mapping table entries can be stored in key-value pair format, where the key is the terminal form encoding field and the value is the target viewing direction field. For example, if the terminal form encoding field is represented by `code`, and `code=1` represents the top installation form and `code=2` represents the bottom installation form, then the mapping table storage area can store at least the following mapping table entries: when `code=1`, it corresponds to the target viewing direction field `target_dir_up`, and when `code=2`, it corresponds to the target viewing direction field `target_dir_down`. The specific expression of the target viewing direction field is not limited; it can be the target gaze point coordinates, target direction parameters, or other equivalent direction descriptions, as long as it can be used for subsequent guidance and comparison.

[0040] During the data entry and authentication phases, the data entry device retrieves the target gaze direction field based on the acquired terminal form encoding field to determine the target gaze direction. To facilitate reproduction by those skilled in the art, the retrieval process can be illustrated as follows: Assuming the mapping table is Map and the terminal form encoding field is code, then the following can be retrieved using the key: target_dir=Map[code] When there is a mapping table entry corresponding to the code in the mapping table, target_dir is the target line-of-sight direction field obtained through retrieval; when there is no mapping table entry corresponding to the code in the mapping table, it can be processed according to a preset strategy, such as prompting to re-obtain the terminal form code, adopting a default target line-of-sight direction, or entering an exception handling process. Through the above settings, the target line-of-sight direction can be consistently determined based on the same terminal form code field during both the input stage and the authentication stage, thereby ensuring that the guiding and authentication processes under different terminal forms have consistent call logics and reproducibility.

[0041] In one embodiment, the line-of-sight direction parameter output by the eye tracking unit includes a line-of-sight direction vector, which is used to represent the direction that the operator is currently looking at; the target line-of-sight direction is characterized by a target line-of-sight direction vector, which is used to represent the target fixation direction expected in the current terminal form. For ease of implementation, both the line-of-sight direction vector and the target line-of-sight direction vector can be represented in the form of three-dimensional vectors and can be normalized to make their lengths equal to 1. For example, they are respectively denoted as: view_vec=(vx,vy,vz) target_vec=(tx,ty,tz) where view_vec is the line-of-sight direction vector and target_vec is the target line-of-sight direction vector.

[0042] In one embodiment, the input device calculates the angle between the line-of-sight direction vector and the target line-of-sight direction vector, and determines whether to trigger iris image acquisition based on the angle. For ease of reproduction, the angle angle_now can be calculated as follows:

[0043] where arccos(.) is the inverse cosine function, and angle_now represents the magnitude of the angle between the two vectors. The input device compares angle_now with a preset angle threshold angle_limit. When angle_now is less than angle_limit, it is determined that the current line-of-sight direction meets the acquisition requirements.

[0044] In one embodiment, to improve the triggering stability, a preset time window time_window is set, and it is required that angle_now < angle_limit be continuously satisfied within the preset time window to drive the iris image acquisition unit to perform iris image acquisition. Taking the output frequency of the eye tracking unit as fps, the required number of consecutive satisfactions count_need can be determined as follows:

[0045] Where ceil(.) is the ceiling function. The triggering strategy can be: whenever a detection satisfies angle_now < angle_limit, the continuous count is incremented by 1; when the continuous count reaches count_need, the iris acquisition imaging unit is driven to acquire an iris image; if any detection does not satisfy angle_now < angle_limit during the process, the continuous count is cleared and the counting restarts, thus avoiding false triggering caused by short-term jitter.

[0046] In one embodiment, the eye movement tracking unit also outputs a blink state parameter, which is used to characterize whether blinking or eye occlusion occurs. For ease of implementation, the blink state parameter can be denoted as blink_flag, where blink_flag = 1 indicates that blinking or occlusion occurs, and blink_flag = 0 indicates that no blinking or occlusion occurs. Within a preset time window, when blink_flag indicates that blinking or eye occlusion occurs, the input device stops driving the iris acquisition imaging unit, clears the continuous count, and simultaneously restarts the timing of the preset time window; after blink_flag returns to the non-blinking / non-occlusion state, the continuous determination of angle_now and angle_limit is restarted. Through the above settings, it is possible to avoid invalid images caused by triggering during blinking or occlusion from entering the subsequent processing flow and improve the effectiveness and stability of iris acquisition.

[0047] In one embodiment, the input device includes a quality evaluation unit, which is used to evaluate the quality of the iris image acquired by the iris acquisition imaging unit, so as to screen out iris images that meet the quality requirements for subsequent template generation, thereby improving the stability and usability of the iris template obtained in the input stage.

[0048] In the input stage, for the same terminal form code, the iris acquisition imaging unit can acquire multiple iris images at different time points, for example, continuously acquire an image sequence I1, I2,..., In at a preset acquisition interval. The quality evaluation unit calculates the quality index for each iris image respectively, and compares the quality index with a preset quality threshold to determine whether the iris image meets the quality requirements.

[0049] For ease of reproduction by those skilled in the art, the quality index can be composed of at least one computable image quality quantization result, or obtained by weighting multiple quality quantization results. For example, it can at least include one or more of a clarity index, an effective iris area ratio index, and a reflection / occlusion index.

[0050] In an alternative embodiment, the clarity index can be calculated based on the image edge information. For example, if the Laplacian operator result of image I is denoted as Lap(I), then the clarity index sharp(I) can be taken as: sharp(I) = Var(Lap(I)) Var(.) represents variance operation; a larger sharp(I) usually indicates a sharper image.

[0051] In one optional implementation, the effective ratio of the iris region can be calculated based on the iris region segmentation results. Let IrisArea(I) represent the set of pixels in the iris region, and OccArea(I) represent the set of pixels in the occluded region (e.g., the area occluded by eyelids, eyelashes, or eyeglass frames), then the effective ratio valid_ratio(I) can be taken as: valid_ratio(I)=1-|OccArea(I)| / |IrisArea(I)| Where |.| represents the number of pixels; the larger the valid_ratio(I), the less occlusion there is.

[0052] In one alternative implementation, the reflectivity index can be calculated based on the proportion of bright pixels. Let GlareArea(I) represent the set of pixels in the reflective bright area, then the reflectivity suppression index glare_score(I) can be taken as: glare_score(I)=1-|GlareArea(I)| / |IrisArea(I)| A higher glare_score(I) indicates less reflection.

[0053] In one implementation, the quality assessment department may combine one or more of the above-mentioned quantitative results to form a total quality index Q(I). For example, a weighted method may be used to obtain:

[0054] w1, w2, and w3 are weighting coefficients, which can be obtained through system configuration; when a certain indicator is not used, its corresponding weight can be set to 0 or not included in the calculation. The preset quality threshold is denoted as Q_limit.

[0055] The quality assessment process can be as follows: Calculate Q(Ij) for each iris image Ij and compare it with a preset quality threshold Q_limit. When Q(Ij) meets the quality threshold requirement (e.g., Q(Ij)>=Q_limit), generate an iris template based on the corresponding iris image Ij, establish a correspondence between the generated iris template and the current terminal morphology code, and store it. When Q(Ij) does not meet the quality threshold requirement, maintain the presentation of the guide point and continue to acquire subsequent iris images to obtain iris images that meet the quality requirements for template generation. Through this setting, blurry, occluded, or highly reflective iris images can be avoided for template generation, thereby improving the quality of iris templates during the data entry stage and the stability of subsequent authentication.

[0056] In one embodiment, the input device or the terminal to be authenticated includes a storage structure for the organized storage and retrieval of iris templates, so that the set of templates to be matched can be quickly located based on the terminal morphology code during the authentication stage. The storage structure includes a template storage area and an index storage area, wherein the template storage area is used to store the iris template itself, and the index storage area is used to store index entries to realize the association between the "terminal morphology code" and the "template set".

[0057] In one implementation, the template storage area stores multiple iris templates, each corresponding to a template identifier field `template_id`, used to uniquely identify the iris template. For ease of implementation, the template storage area can be understood as a storage structure with the template identifier field as the key and the iris template data as the value, for example: TemplateStore[template_id]=template_data Where template_data is the iris template data, which can be a fixed-length feature vector or a bit string, and the specific form is not limited.

[0058] In one embodiment, the index storage area stores index entries, each including at least a terminal morphology encoding field (code) and a template identifier field (template_id). There is a one-to-many correspondence between the terminal morphology encoding field and the template identifier field; that is, the same terminal morphology encoding field (code) can correspond to multiple template identifier fields (template_id), allowing multiple iris templates to be stored under the same morphology to improve robustness. For ease of reproduction, the index storage area can be expressed in the following form: IndexStore[code]={template_id_1,template_id_2,...,template_id_m} Where m is the number of templates corresponding to code, and m can be 1 or greater than 1.

[0059] During the authentication phase, after the data entry device obtains the terminal morphology code field "code" from the terminal to be authenticated, it determines the target iris template set corresponding to the template identifier field in the index storage area based on the terminal morphology code field, and performs matching only on the target iris template set to output the authentication result. For ease of implementation by those skilled in the art, the process of determining the target iris template set can be exemplified as follows: First, the set of template identifier fields corresponding to "code" is retrieved from the index storage area: TargetIDs=IndexStore[code] Subsequently, the corresponding iris template data is read from the template storage area based on the template identifier field set to form the target iris template set: TargetTemplates={TemplateStore[id]foridinTargetIDs} Finally, only the iris templates in TargetTemplates are matched against the iris template to be authenticated to output the authentication result, without matching iris templates corresponding to other terminal morphology codes. This setting limits the matching scope to the set of templates corresponding to the current terminal morphology code, reducing the computational overhead of matching irrelevant templates and lowering the probability of mismatches caused by mixing cross-morphology templates, thereby improving the efficiency and accuracy of the authentication process.

[0060] In one embodiment, during the authentication phase, the data entry device first obtains the terminal morphology code from the terminal to be authenticated, and selects the corresponding iris template set for matching based on the obtained terminal morphology code. When the terminal morphology code output by the terminal to be authenticated is missing or does not meet the preset validity conditions, the data entry device enters the candidate terminal morphology code determination process to ensure that authentication matching can still be completed even if the code cannot be reliably obtained.

[0061] In one implementation, "missing terminal form code or not meeting preset validity conditions" may include, but is not limited to: not receiving data containing the terminal form code field, the terminal form code field being empty, the terminal form code field not being in the preset code set, or inconsistent verification results of the terminal form code field. For ease of reproduction, the terminal form code field can be denoted as `code_read`, and the validity judgment result as `valid`. The process for entering the candidate encoding process can then be determined as follows: valid=(code_readisnotNULL)AND(code_readinValidCodeSet)ANDCheckPass(code_read) Where ValidCodeSet is a preset set of valid codes (e.g., including upper and lower codes), and CheckPass(.) represents the verification and judgment process (e.g., checksum or CRC consistency judgment), the specific verification method is not limited. When valid is false, the data entry device enters the candidate code determination process.

[0062] In the candidate terminal morphology encoding determination process, the input device determines the candidate terminal morphology encoding based on the deviation between the gaze direction parameter and each target gaze direction in the preset mapping relationship. Specifically, the preset mapping relationship contains multiple mapping entries of "terminal morphology encoding field - target gaze direction field". The input device extracts the target gaze direction field for each mapping entry and calculates the deviation between the current gaze direction parameter and the target gaze direction. The smaller the deviation, the more the current gaze state matches the terminal morphology corresponding to that mapping entry.

[0063] For ease of implementation, the view direction vector corresponding to the current view direction parameter can be denoted as `view_vec`, and the target view direction vector corresponding to each mapping table entry can be denoted as `target_vec(code)`, with the included angle used as the deviation. For any candidate code `code_i`, its deviation `angle_i` can be calculated as follows: cos_i=dot(view_vec,target_vec(code_i)) angle_i = arccos(cos_i) Where dot(a,b) represents the vector dot product, and arccos(.) represents the inverse cosine function. Subsequently, the recording device selects the candidate terminal morphology code with the smallest deviation from all candidate codes, for example: code_candidate=argmin(angle_i) In one implementation, to avoid misselection, candidate reliability conditions can be further set. For example, candidate encoding is only used when the minimum deviation meets a preset deviation threshold; if not, the system prompts the user to reacquire the terminal form code or restart the boot process. Examples of candidate reliability conditions are as follows: ifmin(angle_i) <cand_limitthenacceptcode_candidateelsefallback Once the candidate terminal morphology code is determined, the data entry device selects the corresponding iris template or set of iris templates for matching based on that candidate terminal morphology code. For example, the target template identifier set can be retrieved from the index storage area based on the candidate terminal morphology code field, and the corresponding target iris template set can be retrieved from the template storage area. Matching is only performed on this target iris template set to output the authentication result. Through the above settings, even if the terminal to be authenticated fails to output a valid terminal morphology code, the candidate terminal morphology code can still be determined by utilizing the deviation between the gaze direction and the preset mapping relationship, and template selection and matching can be completed accordingly, improving the availability and robustness of the system in cases of abnormal communication or abnormal terminal configuration.

[0064] In one embodiment, the storage structure includes a template storage area, an index storage area, and an update rule storage area. The update rule storage area stores update rules corresponding to the terminal morphology code, so that when the system generates a new iris template for the same terminal morphology code during the input stage, it can write the new template into the template storage area according to a preset strategy and simultaneously maintain the correspondence between the "terminal morphology code field - template identifier field" in the index storage area, thus avoiding inconsistencies between the index and template data.

[0065] In one implementation, the update rule storage area can store update rules in the following form: using the terminal morphology encoding field `code` as the key and the update rule `Rule(code)` as the value, for example: RuleStore[code] = Rule(code) The Rule(code) may include at least the upper limit K(code) for the number of templates to be retained associated with that code, and the template elimination policy Policy(code). K(code) is used to limit the number of templates that can be retained under the same terminal form code, and Policy(code) is used to determine the objects to be eliminated when the number of templates exceeds the upper limit or when replacement is required. K(code) and Policy(code) can be obtained by system configuration, and their specific values ​​are not limited.

[0066] During the data entry phase, when a new iris template is generated for the same terminal morphology encoding, the system assigns a new template identifier field `new_id` to the new iris template, writes the new iris template to the template storage area, and simultaneously updates the index entry in the index storage area corresponding to the terminal morphology encoding field. For ease of reproduction, the following procedure can be followed: (1) Write to template storage area TemplateStore[new_id]=new_template (2) Update index table entries (synchronously write template identifier field) IndexStore[code]=IndexStore[code]U{new_id} Here, TemplateStore represents the template storage area, IndexStore represents the index storage area, and "U" represents the set union operation. Through the above operations, the template identifier field set of the index table entry corresponding to the terminal morphology encoding field code contains new_id, thereby achieving consistency maintenance of "new templates being written to the template storage area and index table entries being updated synchronously".

[0067] In one implementation, when the number of templates in IndexStore[code] exceeds the retention limit K(code) specified by the update rule, eviction is performed according to the update rule, and the template storage area and index storage area are maintained synchronously. For example, if Policy(code) is "eviction of the oldest template", then a write timestamp ts(id) can be maintained for each template identifier field, and the drop_id field of the template to be evicted can be determined: drop_id=argmin(ts(id)),idinIndexStore[code] Then execute: IndexStore[code]=IndexStore[code]\{drop_id} deleteTemplateStore[drop_id] Here, "\" represents the set difference operation. This method ensures that the actual templates in the template storage area are consistent with the template identifier field recorded in the index storage area.

[0068] In another alternative implementation, Policy(code) can also employ a "eliminating the lowest quality template" strategy. In this case, a corresponding quality score Q(id) can be maintained for each template identifier field, and the following can be determined: drop_id=argmin(Q(id)),idinIndexStore[code] The index and template storage are updated synchronously in the same manner as described above, thus prioritizing the retention of higher-quality iris templates. By setting an update rule storage area and writing and synchronously updating index entries according to rules when generating new templates under the same terminal morphology encoding, the template library can be gradually optimized over time, while avoiding disorderly growth in the number of templates or index mismatch, thereby improving the stability and maintainability of subsequent authentication matching.

[0069] In one embodiment, a computer-readable storage medium is provided, wherein computer program instructions are stored in the computer-readable storage medium. The computer-readable storage medium may be a non-volatile storage medium or a removable storage medium, specifically, but not limited to, flash memory, USB flash drive, memory card, solid-state drive, read-only memory, erasable programmable read-only memory, or other media capable of storing computer programs.

[0070] In one embodiment, when the computer program instructions are executed by the processor, the processor performs the iris template entry and authentication process described in the foregoing embodiments. For example, when the processor executes the computer program instructions, it can implement the following functional modules or processes: acquiring and managing terminal morphology codes; establishing and calling the mapping relationship between terminal morphology codes and target gaze directions; during the input stage, presenting guide points based on terminal morphology codes, receiving gaze direction parameters output by eye tracking and comparing them to trigger iris image acquisition; performing quality assessment on the acquired iris images and generating iris templates when quality requirements are met; writing the iris templates into the template storage area and maintaining the index relationship between terminal morphology codes and template identifier fields; during the authentication stage, acquiring terminal morphology codes and determining target gaze directions accordingly, acquiring iris images to be authenticated and generating iris templates to be authenticated when acquisition conditions are met; limiting the matching range based on the index storage area, selecting a set of target iris templates corresponding to the terminal morphology codes and performing matching to output authentication results; when terminal morphology codes are missing or invalid, determining candidate terminal morphology codes based on the deviation between gaze direction parameters and each target gaze direction and selecting iris templates to perform matching accordingly; and synchronously updating and maintaining the template storage area and index storage area according to update rules when generating new iris templates. By embedding the above process in a computer-readable storage medium as a computer program, the relevant device can implement the iris template entry and authentication method described in the foregoing embodiments after loading and executing the computer program.

[0071] In one embodiment, a computer device is provided, the computer device including a memory and a processor. The memory is used to store computer program instructions and data generated during operation, and the processor is used to call and execute the computer program instructions in the memory to implement the iris template enrollment and authentication process described in the foregoing embodiments.

[0072] In one embodiment, the computer device may be a data entry device, a terminal to be authenticated, or any node device in a distributed computing system composed of a data entry device and a terminal to be authenticated. The processor may be a general-purpose processor, a digital signal processor, a microcontroller, or other programmable logic device, and the memory may be a combination of volatile memory and / or non-volatile memory, without limitation on the specific type.

[0073] In one embodiment, when the processor executes the computer program instructions, it can perform the following processing steps: establishing and managing terminal morphology codes and their validity judgment; establishing and storing the mapping relationship between terminal morphology codes and target gaze directions, and retrieving target gaze directions according to terminal morphology codes during the input and authentication phases; presenting guide points and receiving gaze direction parameters output by eye tracking during the input phase, and driving the iris acquisition imaging unit to acquire iris images when the acquisition trigger conditions are met; performing quality assessment on the acquired iris images and generating an iris template when quality requirements are met; writing the iris template into the template storage area and maintaining the terminal morphology in the index storage area. The system establishes a one-to-many correspondence between the encoding field and the template identifier field. During the authentication phase, the system acquires the terminal morphology code and determines the target gaze direction accordingly. When the acquisition trigger condition is met, a template to be authenticated is generated, and a set of target iris templates is determined based on the index storage area. Matching is performed only on the target iris template set to output the authentication result. When the terminal morphology code is missing or invalid, candidate terminal morphology codes are determined based on the deviation between the gaze direction parameters and the target gaze directions in the mapping relationship, and iris templates are selected accordingly for matching. Furthermore, when a new iris template is generated under the same terminal morphology code, the template storage area and the index storage area are synchronously updated and maintained according to update rules. Through these settings, the computer device can implement the iris template input and authentication method described in the preceding embodiments when executing the computer program instructions.

[0074] The embodiments described above are merely examples of several implementations of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A method for iris template input and authentication for multi-form reuse in eye-tracking terminals, characterized in that, Including the following steps: S10. Provide an input device, the input device including an eye-tracking unit and an iris acquisition and imaging unit, and the input device having a gaze interface for presenting guide points; S20. Establish terminal form code and establish a preset mapping relationship between the terminal form code and the target gaze direction. The terminal form code represents the vertical position relationship between the iris acquisition imaging module of the terminal to be authenticated and the gaze interface of the terminal to be authenticated. S30. During the input phase, input operations are performed sequentially for at least two different terminal form codes: the target gaze direction corresponding to the preset mapping relationship is determined based on the current terminal form code, and a guide point is presented on the gaze interface of the input device. The eye-tracking unit outputs the gaze direction parameters and compares them with the target gaze direction. When the allowable deviation is met, the iris acquisition and imaging unit is driven to acquire the iris image and generate the iris template. The iris template is then associated with the current terminal morphology code and stored. S40. During the authentication phase, obtain the terminal form code from the terminal to be authenticated. The eye-tracking unit outputs the gaze direction parameters at the moment of authentication and compares them with the target gaze direction corresponding to the preset mapping relationship. When the allowable deviation is met, the iris template corresponding to the acquired terminal morphology code is selected from the stored iris templates and matched with the iris template generated from the iris image to be authenticated to output the authentication result.

2. The iris template input and authentication method for multi-form reuse in eye-tracking terminals according to claim 1, characterized in that, The terminal to be authenticated includes a gaze interface, an iris acquisition and imaging module, and a control unit electrically connected to the iris acquisition and imaging module. The control unit includes a non-volatile memory and a communication interface. The terminal form code is written into the non-volatile memory, and the terminal form code includes at least an upper form code and a lower form code. The upper form code and the lower form code correspond to the upper and lower installation forms of the iris acquisition imaging module relative to the gaze interface of the terminal to be authenticated, respectively. During the authentication stage, the terminal form code is output through the communication interface.

3. The iris template input and authentication method for multi-form reuse in eye-tracking terminals according to claim 2, characterized in that, The input device includes a mapping table storage area, which stores mapping table entries. Each mapping table entry includes a terminal form encoding field and a target gaze direction field, and the terminal form encoding field corresponds one-to-one with the target gaze direction field. During the input phase and the authentication phase, the target gaze direction field is retrieved based on the terminal form encoding field to determine the target gaze direction.

4. The iris template input and authentication method for multi-form reuse in eye-tracking terminals according to claim 3, characterized in that, The gaze direction parameter output by the eye-tracking unit includes a gaze direction vector, and the target gaze direction is represented by the target gaze direction vector; The input device calculates the angle between the gaze direction vector and the target gaze direction vector, and drives the iris acquisition imaging unit to acquire the iris image when the angle is less than a preset angle threshold and is continuously satisfied within a preset time window. Within the preset time window, when the blink state parameters output by the eye-tracking unit indicate that a blink or eye occlusion has occurred, the driving of the iris acquisition imaging unit is stopped and the timing of the preset time window is restarted.

5. The iris template input and authentication method for multi-form reuse in eye-tracking terminals according to claim 4, characterized in that, The data entry equipment includes a quality assessment department; During the input phase, for the same terminal form encoding, the iris acquisition and imaging unit acquires multiple iris images at different time points; The quality assessment unit obtains quality indicators for each iris image and compares them with a preset quality threshold. When the quality indicators meet the preset quality threshold, the iris template is generated based on the corresponding iris image, and the iris template is associated with the terminal morphology code and stored. If the quality indicators do not meet the preset quality threshold, the presentation of the guide point is maintained and the iris image continues to be acquired.

6. The iris template input and authentication method for multi-form reuse in eye-tracking terminals according to claim 5, characterized in that, The input device or the terminal to be authenticated includes a storage structure, which includes a template storage area and an index storage area. The template storage area stores the iris template; The index storage area stores index entries, each index entry including a terminal morphology encoding field and a template identifier field, with the terminal morphology encoding field and the template identifier field having a one-to-many correspondence; During the authentication phase, the target iris template set corresponding to the template identifier field is determined in the index storage area based on the terminal morphology encoding field, and the matching is performed only on the target iris template set to output the authentication result.

7. The iris template input and authentication method for multi-form reuse in eye-tracking terminals according to claim 6, characterized in that, During the authentication phase, when the terminal form code output by the terminal to be authenticated is missing or does not meet the preset validity conditions, the input device determines the candidate terminal form code based on the deviation between the gaze direction parameter and each target gaze direction in the preset mapping relationship, and selects the iris template to perform the matching according to the candidate terminal form code.

8. The iris template input and authentication method for multi-morphological reuse in eye-tracking terminals according to claim 6 or 7, characterized in that, The storage structure includes an update rule storage area, which stores update rules corresponding to the terminal form code. During the input phase, when a new iris template is generated for the same terminal morphology code, the new iris template is written into the template storage area according to the update rule, and the template identifier field of the index entry corresponding to the terminal morphology code field is updated synchronously.

9. A computer-readable storage medium, characterized in that, The device stores a computer program that, when executed by a processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

10. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

Images