Car owner identity information identification system based on artificial intelligence

Abstract

The invention relates to the technical field of identity recognition, and discloses a vehicle owner identity information recognition system based on artificial intelligence. Comprising a data acquisition module used for acquiring an iris image of a current vehicle operator; the sub-region division module is used for generating an identification region and dividing the identification region into sub-regions; the parameter extraction module is used for generating iris feature coefficients; the comparative analysis module is used for judging whether the vehicle owner identity information exists or not and generating a corresponding vehicle control instruction; the instruction execution module is used for controlling the vehicle to execute corresponding actions according to the vehicle control instruction; according to the invention, an overall identification and judgment mode of the iris image is converted into a local identification and judgment mode, so that the data volume used for vehicle owner identity identification can be reduced, the data calculation workload can be reduced, the control instruction can be generated at the first time, the efficient vehicle owner identity identification effect is realized, and the user experience is improved. And the processing ordering effect of the iris data is also realized.

Description

Technical Field

[0001] This invention relates to the field of identity recognition technology, and more specifically, to an artificial intelligence-based vehicle owner identity information recognition system. Background Technology

[0002] With the rapid development of society and artificial intelligence, accurately identifying the true identity of various personnel has become increasingly important. Iris recognition technology, with its inherent characteristics such as accuracy, non-contact, and liveness detection, overcomes the shortcomings of other biometric technologies while possessing their advantages. It has a high accuracy rate. Therefore, the combination of artificial intelligence and iris recognition technology has also been applied to intelligent vehicles to identify the identity information of personnel entering the vehicle.

[0003] Chinese patent application publication number CN109376725A discloses an identity verification method and device based on iris recognition. Its identity verification platform built on iris recognition technology can accurately identify the real identity of various people, avoiding the security vulnerabilities of single verification methods such as ID card, fingerprint, and face, and eliminating illegal issues such as ID card forgery and impersonation of other people's ID cards. The existing technology has the following shortcomings: When using iris recognition technology to identify personal information, it is necessary to recognize the entire iris. This requires recognizing all the information contained in the iris, resulting in a large amount of data collection and computation. Furthermore, since the recognition result can only be given after the entire iris information is recognized, it is impossible to issue an early warning when abnormal iris information is detected, which further increases the time required for iris recognition. Moreover, the lack of priority distinction in the iris information recognition process leads to disorder in the iris information recognition process.

[0004] In view of this, the present invention proposes an artificial intelligence-based vehicle owner identity information recognition system to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: an artificial intelligence-based vehicle owner identity information recognition system, applied to an in-vehicle computer, comprising: The data acquisition module is used to acquire the iris image of the current vehicle operator within the effective working area; The sub-region segmentation module is used to mark the outer and inner boundary points in the iris image, connect the outer and inner boundary points to generate the recognition region, and divide the recognition region into sub-regions. The parameter extraction module is used to extract the comprehensive feature parameters of the sub-region. The comprehensive feature parameters include the area ratio of pits, iris radius and horizontal and vertical texture ratio. The module performs overlap analysis on the comprehensive feature parameters to generate effective feature values. The comparison and analysis module is used to determine priorities and analyze the effective feature values ​​of sub-regions in the identification area according to the priorities to determine whether there is vehicle owner identity information. The instruction execution module is used to formulate vehicle control instructions and control the vehicle to execute those instructions.

[0006] Furthermore, methods for acquiring iris images include: The camera continuously acquires p real-time images within the effective working area; Load p real-time images into MATLAB, use row and column indices to specify the position of all pixels, and use the indices to retrieve the pixel values ​​of all pixels in the image matrix. Compare and analyze the pixel values ​​of all pixels in p real-time images with preset first and second pixel thresholds, and mark the pixels whose pixel values ​​are between the first and second pixel thresholds as target pixels; The total number of target pixels is accumulated and compared with all pixels to obtain the pixel ratio; Real-time images with a pixel ratio greater than or equal to a preset ratio threshold are labeled as iris images. An iris image.

[0007] Furthermore, methods for generating the recognition region include: The first scanned image is obtained by scanning the iris image with a laser scanner. All target pixels are marked in the first scanned image to obtain the region corresponding to the target pixels. Mark the area outside the region corresponding to the target pixel at equal first preset length intervals. Mark the outer boundary points at equal intervals of the second preset length. One internal boundary point; Will outer boundary points and Connecting the inner boundary points sequentially with curves forms the outer boundary line and the inner boundary line. The area between the outer boundary line and the inner boundary line is the recognition area, thus obtaining... Each identification region.

[0008] Further methods for obtaining the area ratio of the pits include: A second scanned image is obtained by scanning a sub-region using a laser scanner, and target pixels are marked in the second scanned image; The second scanned image is divided into sections according to a preset width and length. Each small square is used to calculate the sub-area corresponding to the target pixel within each small square, and then... The total area of ​​the iris is obtained by accumulating the individual facets. In the second scanned image, the pixel values ​​of the target pixels are marked, and the target pixels with pixel values ​​greater than a preset third pixel threshold are recorded as concave pixels; Calculate the concave area corresponding to each concave pixel in the small square, and then... The area of ​​the pit is obtained by accumulating the concave surfaces; The area of ​​the pit is compared with the total area of ​​the iris to calculate the ratio of the pit area.

[0009] Furthermore, methods for obtaining the iris radius include: The second scanned image is divided into equal angles. Measure sequentially within a radius region. The distance between the outer and inner boundary points of a radius region is denoted as the sub-radius. Remove the maximum and minimum values ​​of the sub-radius, and... The iris radius is obtained by averaging the sum of the individual radii.

[0010] Further methods for obtaining the horizontal and vertical texture ratios include: The Gabor function is used to define a Gabor filter, the imgaborfilt function is used to filter the second scan image, texture is extracted from the filtered second scan image, and lines are drawn at the locations of the extracted textures to obtain texture lines. In the second scan image, respectively perform Construct a coordinate system by extending the lines containing the sub-radius, and marking the intersection of the extensions as the origin; Measure the angle between the texture line and the X-axis of the coordinate system. Texture lines with an angle greater than 45 degrees are marked as vertical textures, and texture lines with an angle less than or equal to 45 degrees are marked as horizontal textures. The number of vertical textures and the number of horizontal textures are accumulated separately and compared to generate the ratio of vertical to horizontal textures.

[0011] Furthermore, the method for generating effective feature values ​​is as follows: When the pit area ratio is between the lower limit and the upper limit of the pit ratio, the pit area ratio is recorded as a valid parameter. When the iris radius is between the lower and upper limits of the radius, the iris radius is recorded as a valid parameter. When the horizontal and vertical texture ratio is between the lower limit and the upper limit of the texture ratio, the horizontal and vertical texture ratio is recorded as a valid parameter. The number of valid parameters in each sub-region is counted to obtain valid feature values.

[0012] Furthermore, the priority is as follows: the sub-regions closer to the upper and lower eyelids have a higher priority than the sub-regions on the left and right sides; one by one The identified regions are arranged into a region queue. Identify them one by one according to priority. Each identification area The size of the effective feature values ​​in each sub-region; when When the effective feature value of each sub-region is 3, it is determined that the vehicle owner's identity information exists in the recognition area. when When the effective feature value of each sub-region is 0, 1, or 2, it is determined that there is no vehicle owner identity information in the recognition area.

[0013] Furthermore, vehicle control commands include safety activation commands and hazard warning commands; The methods for generating safety activation commands and hazard warning commands include: When the presence of vehicle owner identity information is detected, a security unlock command is generated; When it is determined that there is no vehicle owner identity information, a danger warning command is generated.

[0014] Furthermore, when a safe unlocking command is generated, the onboard computer sends an unlocking command to the door electronic lock controller to control the door electronic lock to unlock; When a hazard warning command is generated, the onboard computer sends activation commands to the lighting controller and horn controller, controlling the vehicle to emit flashing lights and horn alarms.

[0015] The technical advantages of the artificial intelligence-based vehicle owner identity information recognition system of this invention are as follows: This invention divides the acquired iris image into recognition regions to obtain the smallest region corresponding to the iris parameters, thereby avoiding interference from irrelevant parameters of the pupil and sclera on the acquisition and calculation of iris parameters. By dividing the recognition region into multiple sub-regions, the comprehensive iris parameters within each sub-region are acquired and calculated sequentially to generate iris feature coefficients. Simultaneously, the obtained iris feature coefficients are compared and analyzed with pre-stored iris feature coefficients. This method breaks down the parameters of a large recognition region into smaller parts. By using local small-scale comparison, the presence of vehicle owner information can be determined immediately upon detecting inconsistencies in the iris feature coefficient comparison results. Based on the determination result, corresponding control commands are formulated. Compared with existing technologies, this method of converting the overall recognition judgment of iris images into a local recognition judgment method can reduce the amount of data used for vehicle owner identification and the workload of data calculation. It can also generate control commands immediately, achieving efficient vehicle owner identification and orderly processing of iris data. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of an artificial intelligence-based vehicle owner identity information recognition system provided in Embodiment 1 of the present invention; Figure 2 A schematic diagram of dividing the identification area into sub-regions is provided for Embodiment 1 of the present invention; Figure 3 This is a flowchart illustrating the iris-based vehicle owner identification method provided in Embodiment 2 of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1: Please refer to Figure 1 and Figure 2 As shown in this embodiment, an artificial intelligence-based vehicle owner identity information recognition system is applied to an in-vehicle computer. The system includes: The data acquisition module is used to acquire the iris image of the current vehicle operator; Iris images refer to images captured by cameras on vehicles that directly display the iris area of ​​the vehicle operator's eyes, and serve as the basis for subsequent identification of the vehicle owner's identity information; before collecting iris images, it is necessary to capture them in conjunction with cameras installed on the vehicle; In this embodiment, when the vehicle operator approaches the vehicle and activates the camera on the driver's side B-pillar, the camera turns on and displays the captured image in real time on the display screen on the B-pillar. The vehicle operator lowers their head and places their eyes within the effective working area of ​​the camera, thus ensuring that the camera captures images of the vehicle operator's eyes.

[0019] It should be noted that the effective working area of ​​a camera is the area within which the camera can capture a complete image of the vehicle operator's eyes. When the vehicle operator's eyes are within the effective working area of ​​the camera, the image of both eyes remains complete and without any gaps, accurately representing the vehicle operator's identity information; conversely, the opposite is true. The effective working area of ​​the camera is not a fixed area and can be set according to the actual vehicle height and the driver's height. When the vehicle operator's eyes are not within the effective working area of ​​the camera, the vehicle operator needs to adjust the height and angle of their eyes to ensure that their eyes are within the effective working area of ​​the camera.

[0020] Methods for acquiring iris images include: The camera continuously acquires p real-time images within the effective working area; Load p real-time images into MATLAB, use row and column indices to specify the position of all pixels, and use the indices to retrieve the pixel values ​​of all pixels in the image matrix. The pixel values ​​of all pixels in p real-time images are compared and analyzed with preset first and second pixel thresholds. Pixels whose pixel values ​​fall between the first and second pixel thresholds are marked as target pixels. The preset first and second pixel thresholds are the minimum and maximum pixel values ​​for determining the pixel values ​​corresponding to the iris in the real-time images. Since the pixels in the real-time images correspond to the iris, pupil, sclera, and environmental background, and the pixel values ​​corresponding to the iris, pupil, sclera, and environmental background (including natural environment, facial skin, and facial hair) are different, it is necessary to set the first and second pixel thresholds in order to distinguish the pixel values ​​corresponding to the iris in the real-time images. Pixels whose pixel values ​​fall between the first and second pixel thresholds are the pixels corresponding to the iris. The specific values ​​of the first and second pixel thresholds are set based on the pixel values ​​of multiple sets of iris pixels collected in history. The total number of target pixels is accumulated and compared with all pixels to obtain the pixel ratio; Real-time images with a pixel ratio greater than or equal to a preset ratio threshold are labeled as iris images. The image contains iris images; the preset ratio threshold is used to determine whether the pixel ratio of the target pixels in the real-time image reaches the minimum pixel ratio of the target pixels in the iris image. When the pixel ratio is greater than or equal to the preset ratio threshold, it means that there are more target pixels in the real-time image, and the real-time image can be used as an iris image. When the pixel ratio is less than the preset ratio threshold, it means that there are fewer target pixels in the real-time image, and the real-time image cannot be used as an iris image. The preset ratio threshold is set based on the values ​​of multiple sets of pixel ratios collected in history.

[0021] The region segmentation module identifies and marks the regions in the iris image sequentially. outer boundary points and The inner boundary points will be marked. outer boundary points and Connect the inner boundary points sequentially to generate the recognition region, and then divide the recognition region into sub-regions; The recognition region is the smallest area that can completely represent the iris of the vehicle operator. There is only one recognition region for an iris image, and one recognition region can contain all the information of the iris in the iris image.

[0022] Methods for generating the recognition region include: The first scanned image is obtained by scanning the iris image with a laser scanner. All target pixels are marked in the first scanned image to obtain the region corresponding to the target pixels. Mark the area outside the region corresponding to the target pixel at equal first preset length intervals. Mark the outer boundary points at equal intervals of the second preset length. There are two inner boundary points; the first preset length and the second preset length are used to limit the numerical distance between two adjacent outer boundary points and two adjacent inner boundary points, respectively, to ensure that two adjacent outer boundary points and two adjacent inner boundary points maintain a fixed interval distance, thereby preventing the points from being too close or even overlapping. Will outer boundary points and Connecting the inner boundary points sequentially with curves forms the outer boundary line and the inner boundary line. The area between the outer boundary line and the inner boundary line is the recognition area, thus obtaining... A recognition region is generated; by generating a recognition region, the smallest area containing all information of the iris can be obtained. At this time, all information in the recognition region is data unique to the iris. On the one hand, the iris information in the recognition region can be directly extracted and recognized in the subsequent process without additional reprocessing of the extracted and recognized information. On the other hand, it can also greatly reduce the amount of information collected and avoid the negative impact of information contained in the pupil, sclera and environmental background on subsequent information processing and calculation, thereby improving the information processing and calculation speed.

[0023] It should be noted that: in ideal circumstances, the recognition area is a circular structure that can completely represent all the information of an iris. However, in reality, due to the obstruction of the upper and lower eyelids, the top and bottom of the recognition area are flush, so the recognition area is an approximately circular structure.

[0024] Sub-regions are smaller regions obtained by breaking down a large recognition region into smaller ones. In other words, a large recognition region containing a lot of data can be divided into sub-regions of equal size containing less data. When dividing the recognition region into sub-regions, the main criterion is that the areas are of equal size. Therefore, the recognition region needs to be divided into... A sub-region with a uniform area.

[0025] For example, such as Figure 2 As shown, The sub-regions are labeled Y1, Y2, Y3, and Y4, with Y1 located near the upper eyelid and Y3 near the lower eyelid.

[0026] The parameter extraction module extracts the comprehensive feature parameters of each sub-region and performs overlap analysis on the comprehensive feature parameters to generate effective feature values. Comprehensive feature parameters refer to parameters that can reflect the distribution, location, and characteristics of the iris. Since each person's iris is unique, the comprehensive feature parameters of the iris of each vehicle operator will not be completely consistent. By collecting the comprehensive feature parameters of the iris of vehicle operators, the basis for identification and judgment of the vehicle operator's identity information can be provided.

[0027] Comprehensive feature parameters include the area ratio of pits, iris radius, and the ratio of horizontal to vertical textures; The pit area ratio is a parameter that reflects the size of the pit area in the iris relative to the total area of ​​the iris. The size of the pit area ratio is different for each iris. Methods for obtaining the area ratio of pits include: A second scanned image is obtained by scanning a sub-region using a laser scanner, and target pixels are marked in the second scanned image; The second scanned image is divided into sections according to a preset width and length. Each small square is used to calculate the sub-area corresponding to the target pixel within each small square, and then... The total area of ​​the iris is obtained by accumulating the individual facets. The expression for the total area of ​​the iris is: ; In the formula, The total area of ​​the iris. For the first Individual area; In the second scanned image, the pixel values ​​of target pixels are marked, and target pixels with pixel values ​​greater than a preset third pixel threshold are recorded as concave pixels. Since the depth of the concave area in the iris is greater than that of other locations, the pixel values ​​of pixels in the concave area will be greater than the pixel values ​​of pixels in other locations. The preset third pixel threshold can distinguish the pixel values ​​of pixels in the concave area from the pixel values ​​of pixels in other locations. That is, when the pixel value of a target pixel is greater than the third pixel threshold, the target pixel is a concave pixel. The specific value of the third pixel threshold is set based on the pixel values ​​of pixels in multiple sets of concave areas collected in history. Calculate the concave area corresponding to each concave pixel in the small square, and then... The area of ​​the pit is obtained by accumulating the concave surfaces; The expression for the area of ​​the pit is: ; In the formula, The area of ​​the pit. For the first The area of ​​the concave section; After comparing the area of ​​the pit with the overall area of ​​the iris, the ratio of the pit area is calculated. The expression for the proportion of the pit area is: ; In the formula, This represents the area ratio of the pit.

[0028] It should be noted that: each small square has a minimum length calculation unit and a minimum width calculation unit. The sub-area and concave area are only calculated when the length and width of the target pixel and the concave pixel reach the minimum calculation unit. This eliminates data that has a negligible impact on the proportion of the concave area and avoids the tedious workload caused by the statistical calculation of a large number of tiny sub-areas and concave areas. Specifically, the minimum length calculation unit and the minimum width calculation unit have the same value, which is one-tenth of the preset length and width of the small square.

[0029] The iris radius is the distance from the outer edge of the iris to the edge of the pupil, and it can be used to distinguish different sizes and shapes of irises; Methods for obtaining iris radius include: The second scanned image is divided into equal angles. Measure sequentially within a radius region. The distance between the outer and inner boundary points of a radius region is denoted as the sub-radius. Remove the maximum and minimum values ​​of the sub-radius, and... The iris radius is obtained by averaging the sum of the individual radii. The expression for the iris radius is: ; In the formula, The radius of the iris. For the first Sub-radius.

[0030] The ratio of horizontal to vertical textures refers to the ratio between the number of horizontal textures and the number of vertical textures in the iris, and is used to distinguish the texture distribution of different irises. Methods for obtaining the horizontal and vertical texture ratios include: The Gabor function is used to define a Gabor filter, the imgaborfilt function is used to filter the second scan image, texture is extracted from the filtered second scan image, and lines are drawn at the locations of the extracted textures to obtain texture lines. In the second scan image, respectively perform Construct a coordinate system by extending the lines containing the sub-radius, and marking the intersection of the extensions as the origin; Measure the angle between the texture line and the X-axis of the coordinate system. Texture lines with an angle greater than 45 degrees are marked as vertical textures, and texture lines with an angle less than or equal to 45 degrees are marked as horizontal textures. The number of vertical textures and the number of horizontal textures are accumulated separately and compared to generate the ratio of vertical to horizontal textures; The expression for the horizontal and vertical texture ratio is: ; In the formula, The ratio of horizontal to vertical textures, The number of horizontal textures, This represents the number of vertical textures.

[0031] The number of parameters with a high degree of matching effectiveness is important. Since each person's iris may change at different times and in different environments, the comprehensive feature parameters of each vehicle operator's iris will not be completely consistent and there will be some deviation. Therefore, it is necessary to analyze the real-time collected comprehensive feature parameters against the corresponding preset safety thresholds to provide a basis for the identification and judgment of the vehicle operator's identity information.

[0032] The method for generating effective eigenvalues ​​is as follows: The proportion of the dent area in the sub-region is compared with the lower limit and upper limit of the dent proportion. The lower limit and upper limit of the dent proportion refer to the minimum and maximum values ​​of the dent area proportion corresponding to the iris of the car owner, respectively, and are used as one of the criteria for determining whether the person belongs to the car owner. Specifically, the lower limit and upper limit of the dent proportion are obtained by averaging the minimum and maximum values ​​of the dent area proportion corresponding to the iris of a large number of historical data. When the pit area ratio is between the lower limit and the upper limit of the pit ratio, it means that the pit area ratio of the sub-region is within a reasonable range. At this time, the pit area ratio of the sub-region is recorded as a valid parameter. The iris radius of the sub-region is compared with the lower and upper limits of the radius. The lower and upper limits of the radius refer to the minimum and maximum values ​​of the iris radius corresponding to the vehicle owner's iris, respectively, and are used as one of the criteria for determining whether the person is the vehicle owner. Specifically, the lower and upper limits of the radius are obtained by averaging the minimum and maximum values ​​of the iris radius corresponding to a large number of historical data. When the iris radius is between the lower and upper limits, it indicates that the iris radius of the sub-region is within a reasonable range, and the iris radius of the sub-region is recorded as a valid parameter. The horizontal and vertical texture ratios of the sub-regions are compared with the lower and upper limits of the texture ratio. The lower and upper limits of the texture ratio refer to the minimum and maximum values ​​of the horizontal and vertical texture ratios corresponding to the iris of the car owner, respectively, and are used as one of the criteria for determining whether the person is the car owner. Specifically, the lower and upper limits of the texture ratios are obtained by averaging the minimum and maximum values ​​of the horizontal and vertical texture ratios corresponding to a large number of historical data belonging to the iris of the car owner. When the horizontal and vertical texture ratio is between the lower limit and the upper limit of the texture ratio, it means that the horizontal and vertical texture ratio of the sub-region is within a reasonable range. At this time, the horizontal and vertical texture ratio of the sub-region is recorded as a valid parameter. The number of valid parameters in each sub-region is counted to obtain valid feature values.

[0033] In this embodiment, the effective feature value ranges from 0 to 3. When the effective feature value is 0, it means that the area ratio of the pit, the iris radius, and the ratio of horizontal and vertical textures in the sub-region do not coincide with the corresponding safe range. When the effective feature value is 3, it means that the area ratio of the pit, the iris radius, and the ratio of horizontal and vertical textures in the sub-region coincide with the corresponding safe range. The larger the effective feature value, the higher the degree of matching overlap between the sub-region and the vehicle owner.

[0034] The comparison and analysis module sets priorities and analyzes the effective feature values ​​of sub-regions of the identification area according to the priorities to determine whether there is vehicle owner identity information. Because each car owner's iris is unique, the proportions of the pit area, iris radius, and horizontal and vertical texture ratios of the car owner's iris must be controlled within a reasonable range. Therefore, the vehicle operator can only be identified as the car owner if the proportions of the pit area, iris radius, and horizontal and vertical texture ratios of the vehicle operator's iris are highly similar to those of the car owner's iris.

[0035] Since the proportion of pit area, iris radius, and horizontal and vertical texture ratio contained in the sub-regions have varying degrees of influence on the determination of whether someone is a car owner, it is necessary to assign a priority to the sub-regions and compare them according to the priority. When the driver's upper and lower eyelids cover more of the iris due to eye fatigue or illness, the iris characteristic coefficient of the sub-regions near the upper and lower eyelids will be most affected. Therefore, the priority is: the sub-regions closer to the upper and lower eyelids have a higher priority than the sub-regions on the left and right sides.

[0036] Methods for determining whether vehicle owner identity information exists include: one by one The identified regions are arranged into a region queue. Identify them one by one according to priority. Each identification area The size of the effective feature values ​​in each sub-region; when When the effective feature value of each sub-region is 3, it indicates that If the proportion of the pit area, iris radius, and horizontal and vertical texture ratio in a sub-region are highly consistent with the proportion of the pit area, iris radius, and horizontal and vertical texture ratio in the vehicle owner's iris, then it is determined that the vehicle owner's identity information exists in the recognition area. When the effective feature value of the sub-region is 0, 1 or 2, it means that the proportion of the pit area, iris radius and horizontal and vertical texture ratio in the sub-region does not fully overlap with the proportion of the pit area, iris radius and horizontal and vertical texture ratio of the vehicle owner's iris. Therefore, it is determined that there is no vehicle owner identity information in the recognition area.

[0037] For example, firstly, the valid feature values ​​of sub-regions Y1 and Y3 are analyzed. If a valid feature value is not 3, the comparison is stopped, and it is determined that there is no vehicle owner identity information. If the valid feature values ​​of sub-regions Y1 and Y3 are both 3, then the valid feature values ​​of sub-regions Y2 and Y4 are analyzed. If a valid feature value is not 3, the comparison is stopped, and it is determined that there is no vehicle owner identity information. If the valid feature values ​​of sub-regions Y2 and Y4 are both 3, then it is determined that there is vehicle owner identity information.

[0038] The instruction execution module generates vehicle control instructions and controls the vehicle to execute these instructions. Vehicle control commands are specific instructions formulated based on the determination of whether or not vehicle owner identity information exists, which can drive the vehicle to issue different corresponding information when vehicle owner identity information exists or not. Specifically, vehicle control commands include safety unlock commands and hazard warning commands. Safety unlock commands are commands issued when the vehicle owner's identity information is available to control the vehicle to be in a normal unlocked state, while hazard warning commands are commands issued when the vehicle owner's identity information is unavailable to control the vehicle to be in a dangerous or abnormal state.

[0039] The methods for formulating safety activation instructions and hazard warning instructions are as follows: When the vehicle owner's identity information exists, and the vehicle operator performing the identity verification is the vehicle owner, a safe unlocking command will be generated. If the vehicle owner's identity information is not available, and the vehicle operator performing the identity verification is not the vehicle owner, a danger warning instruction will be issued.

[0040] After the safety activation command and hazard warning command are formulated, the vehicle control command will be sent to the on-board computer as soon as possible. This allows the on-board computer to drive and control the relevant equipment and components of the vehicle according to the received vehicle control command, thereby accurately executing and responding to the vehicle control command. Specifically, when a security unlocking command is generated, the on-board computer sends unlocking information to the door electronic lock controller. After receiving the unlocking information, the door electronic lock executes the unlocking action, controlling the vehicle's door electronic locks to unlock. When a hazard warning command is generated, the onboard computer does not send an opening command to the door electronic lock controller, but instead sends an opening information to the light controller and horn controller. At this time, the vehicle's lights flash alternately and the horn sounds intermittently, controlling the vehicle to issue a warning with flashing lights and a horn.

[0041] In this embodiment, by dividing the acquired iris image into recognition regions, the smallest region corresponding to the iris parameters can be obtained, thereby avoiding the interference of irrelevant parameters of the pupil and sclera on the acquisition and calculation of iris parameters. By dividing the recognition region into multiple sub-regions, the comprehensive iris parameters within each sub-region are acquired and calculated sequentially to generate iris feature coefficients. Simultaneously, the obtained iris feature coefficients are compared and analyzed with pre-stored iris feature coefficients. This breaks down the parameters of a large recognition region into smaller parts. By using local small-scale comparison, the presence of vehicle owner information can be determined immediately upon discovering inconsistencies in the iris feature coefficient comparison results. Based on the determination results, corresponding control commands can be formulated. Compared with existing technologies, this method of converting the overall recognition judgment of iris images into a local recognition judgment method can reduce the amount of data used for vehicle owner identification and the workload of data calculation. It can also generate control commands immediately, achieving efficient vehicle owner identification and orderly processing of iris data.

[0042] Example 2: Please refer to Figure 3 As shown, parts not described in detail in this embodiment are described in Embodiment 1. This embodiment provides a method for a vehicle owner identity information recognition system based on artificial intelligence, applied to an in-vehicle computer. It is implemented through an artificial intelligence-based vehicle owner identity information recognition system, including: S1: Within the effective working area, acquire the iris image of the current vehicle operator; S2: Mark the outer and inner boundary points in the iris image, connect the outer and inner boundary points to generate the recognition region, and divide the recognition region into sub-regions; S3: Extract the comprehensive feature parameters of the sub-region. The comprehensive feature parameters include the area ratio of pits, iris radius and horizontal and vertical texture ratio. Perform overlap analysis on the comprehensive feature parameters to generate effective feature values. S4: Establish priorities, analyze the effective feature values ​​of sub-regions in the identification area according to the priorities, and determine whether there is vehicle owner identity information; S5: Formulate vehicle control commands and control the vehicle to execute the vehicle control commands.

[0043] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An artificial intelligence-based vehicle owner identity information recognition system, applied to an in-vehicle computer, characterized in that, include: The data acquisition module is used to acquire the iris image of the current vehicle operator within the effective working area; The sub-region segmentation module is used to mark the outer and inner boundary points in the iris image, connect the outer and inner boundary points to generate the recognition region, and divide the recognition region into sub-regions. The parameter extraction module is used to extract the comprehensive feature parameters of the sub-region. The comprehensive feature parameters include the area ratio of pits, iris radius and horizontal and vertical texture ratio. The module performs overlap analysis on the comprehensive feature parameters to generate effective feature values. The comparison and analysis module is used to determine priorities and analyze the effective feature values ​​of sub-regions in the identification area according to the priorities to determine whether there is vehicle owner identity information. The instruction execution module is used to formulate vehicle control instructions and control the vehicle to execute those instructions.

2. The vehicle owner identity information recognition system based on artificial intelligence according to claim 1, characterized in that, Methods for acquiring iris images include: The camera continuously acquires p real-time images within the effective working area; Load p real-time images into MATLAB, use row and column indices to specify the position of all pixels, and use the indices to retrieve the pixel values ​​of all pixels in the image matrix. Compare and analyze the pixel values ​​of all pixels in p real-time images with preset first and second pixel thresholds, and mark the pixels whose pixel values ​​are between the first and second pixel thresholds as target pixels; The total number of target pixels is accumulated and compared with all pixels to obtain the pixel ratio; Real-time images with a pixel ratio greater than or equal to a preset ratio threshold are labeled as iris images. An iris image.

3. The vehicle owner identity information recognition system based on artificial intelligence according to claim 2, characterized in that, Methods for generating the recognition region include: The first scanned image is obtained by scanning the iris image with a laser scanner. All target pixels are marked in the first scanned image to obtain the region corresponding to the target pixels. Mark the area outside the region corresponding to the target pixel at equal first preset length intervals. Mark the outer boundary points at equal intervals of the second preset length. One internal boundary point; Will outer boundary points and Connecting the inner boundary points sequentially with curves forms the outer boundary line and the inner boundary line. The area between the outer boundary line and the inner boundary line is the recognition area, thus obtaining... Each identification region.

4. The vehicle owner identity information recognition system based on artificial intelligence according to claim 3, characterized in that, Methods for obtaining the area ratio of pits include: A second scanned image is obtained by scanning a sub-region using a laser scanner, and target pixels are marked in the second scanned image; The second scanned image is divided into sections according to a preset width and length. Each small square is used to calculate the sub-area corresponding to the target pixel within each small square, and then... The total area of ​​the iris is obtained by accumulating the individual facets. In the second scanned image, the pixel values ​​of the target pixels are marked, and the target pixels with pixel values ​​greater than a preset third pixel threshold are recorded as concave pixels; Calculate the concave area corresponding to each concave pixel in the small square, and then... The area of ​​the pit is obtained by accumulating the concave surfaces; The area of ​​the pit is compared with the total area of ​​the iris to calculate the ratio of the pit area.

5. The vehicle owner identity information recognition system based on artificial intelligence according to claim 4, characterized in that, Methods for obtaining iris radius include: The second scanned image is divided into equal angles. Measure sequentially within a radius region. The distance between the outer and inner boundary points of a radius region is denoted as the sub-radius. Remove the maximum and minimum values ​​of the sub-radius, and... The iris radius is obtained by averaging the sum of the individual radii.

6. The vehicle owner identity information recognition system based on artificial intelligence according to claim 5, characterized in that, Methods for obtaining the horizontal and vertical texture ratios include: The Gabor function is used to define a Gabor filter, the imgaborfilt function is used to filter the second scan image, texture is extracted from the filtered second scan image, and lines are drawn at the locations of the extracted textures to obtain texture lines. In the second scan image, respectively perform Construct a coordinate system by extending the lines containing the individual radii and setting the intersection of the extensions as the origin. Measure the angle between the texture line and the X-axis of the coordinate system. Texture lines with an angle greater than 45 degrees are marked as vertical textures, and texture lines with an angle less than or equal to 45 degrees are marked as horizontal textures. The number of vertical textures and the number of horizontal textures are accumulated separately and compared to generate the ratio of vertical to horizontal textures.

7. The vehicle owner identity information recognition system based on artificial intelligence according to claim 6, characterized in that, The method for generating effective eigenvalues ​​is as follows: When the pit area ratio is between the lower limit and the upper limit of the pit ratio, the pit area ratio is recorded as a valid parameter. When the iris radius is between the lower and upper limits of the radius, the iris radius is recorded as a valid parameter. When the horizontal and vertical texture ratio is between the lower limit and the upper limit of the texture ratio, the horizontal and vertical texture ratio is recorded as a valid parameter. The number of valid parameters in each sub-region is counted to obtain valid feature values.

8. The vehicle owner identity information recognition system based on artificial intelligence according to claim 7, characterized in that, The priority is as follows: the sub-regions closer to the upper and lower eyelids have a higher priority than the sub-regions on the left and right sides; one by one The identified regions are arranged into a region queue. Identify them one by one according to priority. Each identification area The size of the effective feature values ​​in each sub-region; when When the effective feature value of each sub-region is 3, it is determined that the vehicle owner's identity information exists in the recognition area. when When the effective feature value of each sub-region is 0, 1, or 2, it is determined that there is no vehicle owner identity information in the recognition area.

9. A vehicle owner identity information recognition system based on artificial intelligence according to claim 8, characterized in that, Vehicle control commands include safety activation commands and hazard warning commands; The methods for generating safety activation commands and hazard warning commands include: When the vehicle owner's identity information is available, a security unlock command is generated; When the vehicle owner's identity information is not available, a danger warning command is generated.

10. A vehicle owner identity information recognition system based on artificial intelligence according to claim 9, characterized in that, When a safe unlocking command is generated, the onboard computer sends an unlocking command to the door electronic lock controller to control the door electronic lock to unlock; When a hazard warning command is generated, the onboard computer sends activation commands to the lighting controller and horn controller, controlling the vehicle to emit flashing lights and horn warnings.

Images