High-precision multispectral iris image acquisition system

Abstract

The invention relates to the technical field of image processing, in particular to a high-precision multispectral iris image acquisition system, which comprises an iris region segmentation module for acquiring an iris image from acquired face information, analyzing local textural features of the iris, identifying different structures in the iris, dividing regions of the image according to the structures and the textural features, and acquiring a multi-spectral iris image; and generating iris local structure region data. According to the method, the iris image is ensured to be divided into a plurality of areas with significant texture differences by acquiring the iris image from the acquired face information, extracting the image structure features and analyzing the local texture features of the iris. By extracting the spectrum change curve, the spectrum characteristic of the area can be identified, and the flexibility of the system in processing complex spectrum information is enhanced. By calculating the light delay of each wave band and adjusting the phase difference, the synchronism of images of each wave band is ensured, the problem of detail dislocation of images of different wave bands is solved, and the integration precision of the images under different spectrums is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a high-precision multi-spectral iris image acquisition system. Background Art

[0002] The field of image processing technology involves the use of computers to process, analyze and understand image data to achieve functions such as enhancement, recognition, classification and extraction. By analyzing image features such as color, brightness, texture, shape, etc., image processing technology can be used for applications such as face recognition, target detection, medical image analysis, remote sensing image processing, and visual systems in autonomous driving.

[0003] Among them, the multispectral iris image acquisition system is used to collect iris images in different spectral ranges to improve the accuracy of iris recognition. By capturing iris images in multiple bands such as visible light and infrared light, it can provide richer iris details than a single spectrum, improving the robustness and accuracy of recognition. This technology is widely used in identity authentication devices such as access control machines and attendance machines to ensure that the system can complete identity recognition in different environments.

[0004] Although the existing technology can capture the basic features of the iris when processing single-spectrum iris images, it is difficult to effectively handle situations with complex lighting or large environmental changes, resulting in insufficient robustness of image recognition. Single-spectrum images are prone to loss of details or insufficient image information when facing different lighting conditions, especially when the light is weak or the reflection is uneven, making it difficult to effectively identify the texture and edge features in the image, affecting the accuracy of identity verification. In addition, the existing technology has difficulty in focusing resources on key areas during the image processing process, resulting in limitations in the overall recognition performance. In addition, in complex environments, iris image processing only performs contrast or brightness adjustments, which is difficult to meet high-demand identity verification scenarios. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a high-precision multi-spectral iris image acquisition system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a high-precision multi-spectral iris image acquisition system, the system comprising:

[0007] The iris region segmentation module obtains the iris image from the collected face information, analyzes the local texture features of the iris, identifies different structures in the iris, divides the image into regions according to the structural and texture features, and generates iris local structure region data;

[0008] The spectral response analysis module analyzes the spectral characteristics of each local area according to the iris local structure area data, compares the spectral response of the bands, extracts the spectral change curve of the local area in each band, and generates the spectral response characteristic data of the area;

[0009] The phase difference compensation module analyzes the phase difference trend of the band according to the spectral response characteristic data of the region, adjusts the delay amount of the band, synchronizes the image details, and generates phase-synchronized image data;

[0010] The spatial entropy enhancement module calculates the entropy value of each band in the image according to the phase-synchronized image data, compares the entropy value distribution, adjusts the brightness and contrast of the selected area, and generates entropy enhanced area image data;

[0011] The feature enhancement module extracts features from the entropy enhanced region image data, identifies and analyzes local edge and texture features, integrates the extracted features, and generates iris image feature integration data;

[0012] The image detail optimization module integrates data according to the iris image features, optimizes and analyzes the edge and texture details of the local area, adjusts the details of the local area, and generates optimized iris image data.

[0013] The present invention has the following improvements: the iris local structure area data includes the texture boundary coordinates of the area, the grayscale distribution of the detail area and the pixel structure information of each area; the spectral response characteristic data of the area includes the spectral intensity under multiple bands, the spectral absorption peak information of the local area and the spectral reflectance ratio between the bands; the phase synchronization image data includes the aligned image pixel position information, the phase adjustment coefficient after band synchronization and the synchronization state mark of the area; the entropy enhancement area image data includes the spatial entropy value distribution of the enhanced area, the adjusted contrast level and the detail enhancement parameters of the area; the iris image feature data includes the edge contour coordinates, the detail change rate of the local texture and the edge feature comparison results under multiple spectra; the optimized iris image data includes the optimized edge sharpening degree, texture consistency parameter and regional detail matching score.

[0014] The present invention is improved in that the iris region segmentation module comprises:

[0015] The structural feature extraction submodule identifies the area of ​​structural changes in the iris image and extracts the key structural layers in the image through boundary analysis. The key structural layers are the information of iris texture, iris edge and pupil and iris junction, and generates iris structural feature data;

[0016] The local segmentation processing submodule segments the iris image into multiple regions based on the iris structural feature data and with reference to the identified structural boundaries, and adjusts the segmentation boundaries by comparing the structural changes within the regions to generate iris local segmentation data;

[0017] The regional texture analysis submodule analyzes the distribution of texture features in the region based on the iris local segmentation data, identifies the texture change trend between regions, compares the texture difference of each region, and generates iris local structure region data.

[0018] The present invention is improved in that the spectral response analysis module comprises:

[0019] The spectral reflection information acquisition submodule acquires the reflection information of the local area under multiple spectral bands based on the iris local structure area data, analyzes the reflection characteristics of the area under each band, extracts the reflection intensity information of the local area, and generates spectral reflection data;

[0020] The spectral absorption information analysis submodule compares the reflection intensity of the region under multiple bands based on the spectral reflection data, calculates the spectral absorption value of the local area, analyzes the absorption change trend under the current band, and generates spectral absorption data;

[0021] The spectral change analysis submodule extracts the spectral change trend between bands based on the spectral absorption data, identifies the spectral characteristics of the region by comparing the spectral response differences in the local area, and generates the spectral response characteristic data of the region.

[0022] The present invention is improved in that the phase difference compensation module comprises:

[0023] The light delay measurement submodule analyzes the light propagation path of the band based on the spectral response characteristic data of the area, measures the delay time of each band of light on the iris surface and each area, compares the light delay changes of the band to extract the delay time difference of each band, and generates the light delay data of the band;

[0024] The phase difference calculation submodule calculates the phase difference between adjacent bands based on the light delay data of the bands and the delay difference information of the bands, analyzes the change trend of the phase difference between the bands, and generates the band phase difference data;

[0025] The detail synchronization processing submodule performs detail synchronization processing based on the band phase difference data by adjusting the phase difference value of the iris image in each band to generate phase synchronization image data.

[0026] The present invention is improved in that the spatial entropy enhancement module comprises:

[0027] The entropy value calculation submodule performs information analysis on the local area of ​​each band iris image based on the phase synchronization image data, calculates the entropy value of the area in the image, identifies the entropy value distribution of each band image, and generates band entropy value distribution data;

[0028] The high entropy region identification submodule extracts the region with information content based on the band entropy value distribution data by comparing the entropy value distribution, analyzes the boundary and position of the selected entropy value region, and generates the selected entropy value region data;

[0029] The detail enhancement processing submodule optimizes the detail information of the region by adjusting the brightness and contrast based on the selected entropy value region data to generate entropy enhanced region image data.

[0030] The present invention is improved in that the feature enhancement module comprises:

[0031] The edge information extraction submodule extracts the edge of the iris image of the entropy enhanced region image data, analyzes the continuity of the edge of the iris image, identifies the edge boundary through the brightness difference of the edge, extracts the starting and ending positions of the edge, and generates edge information data;

[0032] The texture feature recognition submodule compares the texture changes in the region based on the edge information data, identifies the direction and density of the texture under multiple bands, and generates texture feature data;

[0033] The multi-band feature fusion submodule integrates the feature information based on the texture feature data to generate iris image feature integration data.

[0034] The present invention is improved in that the image detail optimization module comprises:

[0035] The edge optimization analysis submodule integrates data based on the iris image features, analyzes the edge contour in the region and compares it with the boundary of the adjacent region, adjusts the edge sharpening strength, and generates edge optimization data;

[0036] The texture distribution optimization submodule analyzes the difference between the texture details in the region and the texture distribution in the surrounding region based on the edge optimization data, adjusts the contrast and consistency of the texture, and generates texture optimization data;

[0037] The regional detail adjustment submodule analyzes the matching degree of the edge and texture features in the region based on the texture optimization data, performs synchronous optimization of the edge and texture, and generates optimized iris image data.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are:

[0039] In the present invention, by obtaining an iris image from the collected facial information, extracting image structural features and analyzing iris local texture features, it is ensured that the iris image can be divided into multiple areas with significant texture differences. By extracting the spectral change curve, the spectral characteristics of the area can be identified, which enhances the flexibility of the system in processing complex spectral information. By calculating the light delay of each band and adjusting the phase difference, the synchronization of the images of each band is guaranteed, the problem of misalignment of image details in different bands is solved, and the integration accuracy of the image under different spectra is improved. By calculating and distributing the entropy value of the image, it is possible to identify high-information areas, and based on this, the brightness and contrast are optimized, which significantly improves the expression of image details. In addition, by extracting, integrating and optimizing the features of edges and textures, the high consistency and clarity of the texture details of the iris image in multiple bands are ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The module diagram of the high-precision multi-spectral iris image acquisition system proposed by the present invention;

[0041] Figure 2 The system framework diagram of the high-precision multi-spectral iris image acquisition system proposed by the present invention;

[0042] Figure 3 This is a schematic diagram of an iris region segmentation module of the present invention;

[0043] Figure 4 This is a schematic diagram of a spectral response analysis module of the present invention;

[0044] Figure 5 This is a schematic diagram of a phase difference compensation module of the present invention;

[0045] Figure 6 This is a schematic diagram of a spatial entropy enhancement module of the present invention;

[0046] Figure 7 It is a schematic diagram of a feature enhancement module of the present invention;

[0047] Figure 8 Schematic diagram of the image detail optimization module of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0049] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0050] See also Figure 1 The present invention provides a technical solution: a high-precision multi-spectral iris image acquisition system, the system comprising:

[0051] The iris region segmentation module obtains the iris image from the collected face information, extracts the structural features of the image, analyzes the local texture features of the iris, combines the regional texture comparison, identifies the different structures in the iris, divides the image into regions according to the structural and texture features, ensures that the texture differences between regions are obvious, and generates the iris local structure region data;

[0052] The spectral response analysis module analyzes the spectral characteristics of each local area based on the local structural area data of the iris, obtains the reflection and absorption data of the area under multi-spectral bands, compares the spectral response of the bands, extracts the spectral change curve of the local area in each band, so as to identify the spectral characteristics of the area and generate the spectral response characteristic data of the area;

[0053] The phase difference compensation module measures the light delay of each band according to the spectral response characteristic data of the region, calculates the phase difference value between adjacent bands, analyzes the phase difference trend of the bands, adjusts the delay amount of the bands, synchronizes the details of the image, realizes the synchronization processing of the details of each band, and generates phase-synchronized image data;

[0054] The spatial entropy enhancement module calculates the entropy value of each band in the image based on the phase-synchronized image data, identifies the high-entropy area, compares the entropy value distribution, builds an entropy value distribution model, sets the high-information area as the enhancement target, adjusts the brightness and contrast of the selected area, improves the details and texture of the area, and generates entropy-enhanced area image data;

[0055] The feature enhancement module extracts features from the image data in the entropy enhancement area, including extracting the edge and texture details of the high entropy area, combining the gradient information of the band, identifying and analyzing the local edge and texture features, integrating the extracted features, and generating iris image feature integration data;

[0056] The image detail optimization module integrates data based on iris image features, optimizes and analyzes the edges and texture details of local areas, combines the overall texture distribution of feature data, compares the feature matching of different areas, adjusts the details of local areas, and generates optimized iris image data.

[0057] The data of the local structure area of ​​the iris include the texture boundary coordinates of the area, the grayscale level distribution of the detail area and the pixel structure information of each area. The spectral response characteristic data of the area include the spectral intensity under multiple bands, the spectral absorption peak information of the local area and the spectral reflectance ratio between bands. The phase synchronization image data include the aligned image pixel position information, the phase adjustment coefficient after band synchronization and the synchronization state identification of the area. The entropy enhanced area image data include the spatial entropy value distribution of the enhanced area, the adjusted contrast level and the detail enhancement parameters of the area. The iris image feature data include the edge contour coordinates, the detail change rate of the local texture and the edge feature comparison results under multiple spectra. The optimized iris image data include the optimized edge sharpening degree, texture consistency parameters and regional detail matching score.

[0058] See also Figure 2 and Figure 3 , the iris region segmentation module includes:

[0059] The structural feature extraction submodule identifies the area of ​​structural changes in the iris image and extracts the key structural layers in the image through boundary analysis. The key structural layers are the information of iris texture, iris edge and pupil and iris junction, and generates iris structural feature data;

[0060] First, it is necessary to scan the overall structure of the iris image, extract the areas with large brightness differences in the image, perform refinement according to boundary changes, identify significant changes in local structure, extract data information on edges, textures, and pupil junctions, use grayscale contrast methods to determine the starting and ending points of the iris texture edge, and locate the edge at the junction of the pupil and iris through continuous gradient analysis, extract the boundary coordinates of the iris texture from the inside to the outside, integrate the significant structural changes in each area into data, and finally sort out the main boundaries and regional morphologies according to the structural level changes of each iris area, summarize the different structural levels, refine the morphological features in the area through boundary information, and generate iris structural feature data.

[0061] The local segmentation processing submodule divides the iris image into multiple regions based on the iris structural feature data and with reference to the identified structural boundaries, and adjusts the segmentation boundaries by comparing the structural changes within the regions to generate iris local division data;

[0062] First, the initial boundary of each local area is determined based on the data of the structural hierarchy. Then, a comparative analysis is performed between regions to extract the boundary points of each region. The texture structure inside the region is determined by grayscale changes, and the structural features within the region are determined using geometric morphological analysis methods. Finally, the boundary position of the segmentation is adjusted according to the structural changes. Different local structures are tracked through the dynamic changes of the boundary position. The changes in the segmented area are compared with those of the initial area. The boundary details are adjusted to ensure that the boundaries between each area and the adjacent area are clear. The structural information of each segmented area is organized into iris local segmentation data to generate iris local segmentation data.

[0063] The regional texture analysis submodule analyzes the distribution of texture features in the region based on the local segmentation data of the iris, identifies the texture change trend between regions, compares the texture differences of each region, and generates the local structural region data of the iris;

[0064] First, the texture data information of each segmented area is classified and processed. The directionality of the texture is determined by analyzing the grayscale value of the texture. The correlation between grayscale distribution and gradient direction change is used to extract the subtle texture in the area. The texture change pattern in the area is compared with the adjacent area. The texture density information in each area is obtained by analyzing the texture contrast of the adjacent areas. Finally, the texture features of each segmented area are integrated and analyzed, and the texture differences of each area are compared. According to the texture change trend in the area, the texture features of different areas are classified to form a complete texture change distribution map, and the local structure area data of the iris is generated.

[0065] See also Figure 2 and Figure 4 , the spectral response analysis module includes:

[0066] The spectral reflection information acquisition submodule obtains the reflection information of the local area under multiple spectral bands based on the local structure area data of the iris, analyzes the reflection characteristics of the area under each band, extracts the reflection intensity information of the local area, and generates spectral reflection data;

[0067] First, the reflectance characteristics of each area are measured according to the preset band range, and the spectral reflectance detection instrument is used to capture the reflectance information in different bands. The reflectance difference between the bands is compared, and the reflection intensity of each area in each band is recorded by optical detection equipment. The reflection intensity data is saved as a spectral reflectance information set. The reflection data in each band is then classified and formatted. The data characteristics of each band are combined to ensure data consistency, and the reflection information of different bands is sorted and integrated. Finally, the spectral reflectance information of each area is output and stored as spectral reflectance data.

[0068] The spectral absorption information analysis submodule compares the reflection intensity of the region in multiple bands based on the spectral reflection data, calculates the spectral absorption value of the local area, analyzes the absorption change trend in the current band, and generates spectral absorption data;

[0069] The reflection information of the region is used to compare the absorption characteristics in different bands. Referring to the regional reflection data, the formula is: Calculate the spectral absorption value of the local area. Wherein, A is the spectral absorption value of the area, which indicates how much incident light intensity the area absorbs, and is obtained by calculating the ratio of the light reflection intensity to the incident light intensity in each band. This value is calculated through spectral reflection data. R is the reflection intensity of the area, which indicates the light intensity reflected by the area in a specific spectral band. It is specifically obtained through a spectral reflection detection instrument and stored through a data recording system. I is the intensity of the incident light, which refers to the light intensity irradiated on the surface of the area in a specific spectral band. This value can be obtained through the set intensity of the light source or an illuminance meter, and is usually calibrated as a preset experimental condition.

[0070] The intensity I of the incident light is obtained from the light source device and set to 1. The reflection intensity R of different areas is obtained through the spectral reflection detection device and set to 0.6. Substitute it into the formula:

[0071]

[0072] The result indicates that the region absorbs 40% of the incident light in this band. This absorption value provides the basis for subsequent analysis. The spectral characteristics of the region are analyzed using the absorption change trend of different bands. First, the absorption value in each band is obtained, and the absorption curve is plotted with the wavelength as the horizontal axis and the absorption value as the vertical axis. By comparing the slope and shape of the absorption curve of the region in different bands, the absorption change trend can be identified. By comparing these change trends, the absorption characteristics of the local area in different spectral bands can be determined. For example, a certain area may show a higher absorption peak in a specific band, indicating that the area has a stronger absorption capacity for light in this band. Combining the absorption trends of multiple regions, the band of the absorption peak and its change trend are analyzed to generate the spectral absorption data of the region.

[0073] The spectral change analysis submodule extracts the spectral change trend between bands based on the spectral absorption data, identifies the spectral characteristics of the region by comparing the spectral response differences in the local area, and generates the spectral response characteristic data of the region;

[0074] First, the spectral absorption information in each band is classified and organized. By recording the changing trend of the absorption value, the areas with large changes in absorption value in each band are identified. The absorption difference information between these bands is integrated, and the absorption values ​​in each band are compared. The absorption trend under band changes is analyzed, and the curve of absorption value changing with band changes is recorded as spectral response information. Then, by comparing the spectral response characteristics between regions, the characteristic differences of each region in each band are extracted, the bands with the most significant changes in the local area are identified, and the spectral characteristics of the region are judged through the response differences of these bands. Finally, the spectral response information of each region is integrated to generate the spectral response characteristic data of the region.

[0075] See also Figure 2 and Figure 5 , the phase difference compensation module includes:

[0076] The light delay measurement submodule analyzes the light propagation path of the band based on the spectral response characteristic data of the region, measures the delay time of each band of light on the iris surface and each region, compares the light delay changes of the band to extract the delay time difference of each band, and generates the band light delay data;

[0077] The light delay in different areas of the iris is measured by multi-spectral detection equipment, and the light delay data of each area is obtained using a high-precision spectrum analyzer. The incident angle, band and regional characteristics of the light when it is incident on the iris surface are recorded together. The light delay time between bands is compared to extract the delay changes in different areas. The data is classified and counted for the light delay time changes in each band. Detailed light propagation delay data is generated through correlation analysis between the delay change and the band characteristics. Finally, combined with the optical characteristics of the band and the region, the data is uniformly output and stored as band light delay data.

[0078] The phase difference calculation submodule calculates the phase difference between adjacent bands based on the band light delay data and the delay difference information of the bands, analyzes the changing trend of the phase difference between the bands, and generates the band phase difference data;

[0079] According to the delay difference information of the band, according to the formula:

[0080]

[0081] Calculate the phase difference Δφ between adjacent bands;

[0082] In the formula, Δt represents the light delay time difference between bands, which is obtained by measuring the light propagation delay time of each band through multi-spectral detection equipment and taking the difference. T represents the period of the light wave, which is calculated by the light wave frequency. f is the frequency of the light wave, which is usually a known value. λ1 and λ2 represent the central wavelengths of two adjacent bands, respectively, which are measured by a spectrometer. The wavelength parameter is used to analyze the influence of the wavelength difference between adjacent bands on the phase difference.

[0083] If the delay time difference of light in adjacent bands is Δt = 0.002 seconds, the light wave period T = 0.01 seconds, and the wavelengths of adjacent bands are λ1 = 600nm and λ2 = 620nm, substitute into the formula:

[0084]

[0085] The results show that the phase difference between adjacent bands is 1.2565 radians. Assuming that the historical phase difference value is 1.25 radians, compared with the historical phase difference value, the results show that the phase difference of this band is small, which means that the light propagation in this band is relatively stable and suitable for synchronization processing of multi-band images to generate band phase difference data.

[0086] The detail synchronization processing submodule performs synchronization processing of details based on the band phase difference data by adjusting the phase difference value of the iris image of each band to generate phase synchronization image data;

[0087] First, the phase difference data of each band is read, and the image processing equipment is used to analyze the phase changes between the bands. The synchronization errors of the images of each band are determined by detecting the edge contours and texture details of the image. The pixel position of the image is adjusted for the bands with large phase differences, and the detail areas are aligned. The details of different bands are accurately aligned through operations such as scaling, translation, and rotation to ensure the synchronization of the images of each band at the same time point. Subsequently, the processed band images are fine-tuned, and secondary corrections are made to each area by comparing the detail changes of the images of different bands to generate phase-synchronized image data.

[0088] See also Figure 2 and Figure 6 , the spatial entropy enhancement module includes:

[0089] The entropy value calculation submodule performs information analysis on the local area of ​​each band iris image based on the phase synchronization image data, calculates the entropy value of the area in the image, identifies the entropy value distribution of each band image, and generates band entropy value distribution data;

[0090] Refer to the local area information of the reference image, according to the formula:

[0091]

[0092] Calculate the entropy value H(X) of the local area;

[0093] In the formula, the higher the entropy value of H(X), the more information is contained in the region, and P(xi ) means the pixel intensity is x i The probability of , usually calculated by the grayscale histogram statistics of the image, log2P(x i ) is the logarithm of the intensity probability, which is used to calculate the entropy value, m is the total number of pixels to be analyzed, Δx j is the position change of pixel j, which is obtained by comparing and analyzing the displacement information of the pixel. j It represents the distance from pixel j to the center of the image. As a weighted term, it ensures that pixels farther from the center have a smaller weight in the entropy calculation. Z is a normalization constant that ensures the comparability of entropy values ​​between different image regions. It is usually obtained through a normalization method.

[0094] If the probability distribution of pixel intensity in a certain area is P(x1)=0.2, P(x2)=0.8, the pixel position change Δx1=5, the distance is d1=10, and the normalization constant Z=1, substitute into the formula:

[0095]

[0096] The result shows that the entropy value of this area is 0.7719 bits. If the historical value is 0.75, the entropy value of this area is slightly higher than the historical value, indicating that this area contains rich texture and detail information, and the band entropy value distribution data is generated.

[0097] The high entropy region identification submodule is based on the band entropy value distribution data. By comparing the entropy value distribution, the region with information content is extracted. The region with information content is the region with high entropy value and rich texture and detail features. The boundary and position of the selected entropy value region are analyzed to generate the selected entropy value region data.

[0098] First, the pixel intensity distribution of each area in the image is analyzed by histogram, and then the entropy value of each area is screened according to the entropy distribution data, and the areas with higher entropy values ​​are extracted. The boundaries of these areas are identified using edge detection technology, and then the texture and detail features in these areas are analyzed. By comparing the entropy distribution between each band, the main texture change characteristics in these high entropy areas are further confirmed, and these areas are classified according to their spatial positions. The boundary outline and corresponding texture information of each area are recorded to ensure that the distribution position and size of each high entropy area are completely extracted and output, and the selected entropy area data is generated.

[0099] The detail enhancement processing submodule optimizes the detail information of the region based on the selected entropy value region data by adjusting the brightness and contrast to generate entropy enhanced region image data;

[0100] First, the average brightness value of each area is analyzed, and the brightness enhancement technology is used to improve the brightness level of the high entropy area. The contrast values ​​of each area are compared, and the local contrast is improved through partition adjustment to enhance the visual effect. Then, the texture details are optimized in the high entropy area. By improving the edge clarity and texture layering, the details in the high entropy area are made more prominent and the expression effect of the details is enhanced. Finally, the enhanced regional image data is integrated to ensure that the details of the overall image are effectively improved, and the entropy enhanced regional image data is generated.

[0101] See also Figure 2 and Figure 7 , the feature enhancement module includes:

[0102] The edge information extraction submodule extracts the edge of the iris image of the entropy enhanced region image data, analyzes the continuity of the edge of the iris image, identifies the edge boundary through the brightness difference of the edge, extracts the starting and ending positions of the edge, and generates edge information data;

[0103] First, the edge characteristics in the high entropy area image are analyzed. By comparing the distribution of image brightness and using the brightness gradient change method, the areas in the image that may have significant edges are detected. These areas are then scanned pixel by pixel, and the edge detection method (such as the Sobel operator) is used to accurately identify the starting and ending points of the edges to ensure that the extracted edges have high continuity and integrity. The edge boundaries are then further refined based on the brightness differences of the edges, and the spatial coordinates and grayscale values ​​of each edge point are recorded to ensure the transition details of the edges in different brightness areas. Finally, all the extracted edge information is classified and sorted, and the integrated edge information data is output.

[0104] The texture feature recognition submodule compares the texture changes in the area based on the edge information data, identifies the direction and density of the texture under multi-band conditions, and generates texture feature data;

[0105] First, for the extracted edge information, the texture distribution around it is analyzed, and texture analysis technology (such as gray-level co-occurrence matrix) is used to detect the texture direction, density and spacing in the area. Then, the texture features of iris images in different bands are compared to identify the texture change trends in different bands, especially the texture change areas near the edges, to identify the detailed features of the texture. Finally, the texture direction and density differences in different bands are compared, the texture change information in each area is classified and summarized, and texture feature data is generated. These data will serve as the basis for multi-band feature fusion.

[0106] The multi-band feature fusion submodule integrates the feature information based on the texture feature data to generate iris image feature integration data;

[0107] First, the texture data of each band is uniformly formatted, and the texture features in different bands are compared. Texture fusion technology (such as the feature fusion method based on PCA) is used to integrate the texture information in each band. Then, the texture direction, density and edge features of each band are analyzed. The texture features in multi-band images are fused through feature similarity matching. The texture and edge features are uniformly classified and integrated to ensure that the feature information of all bands is fully processed, forming a complete feature fusion result and generating iris image feature integration data.

[0108] See also Figure 2 and Figure 8 , the image detail optimization module includes:

[0109] The edge optimization analysis submodule integrates data based on iris image features, analyzes the edge contours within the region and compares them with the boundaries of adjacent regions, adjusts the edge sharpening intensity, and generates edge optimization data;

[0110] First, by analyzing the continuity of the edge contour in the area, the changes between the edge and the boundary of the adjacent area are identified. By comparing the sharpness of the edge, image processing methods (such as gradient calculation) are used to compare the brightness changes of the edge and the background area, and the smoothness of the edge transition is evaluated. For the parts with blurred edges or unclear transitions, the edge sharpening intensity is gradually adjusted to make the boundary clearer. Subsequently, multiple iterations are performed within different sharpening intensity ranges. By comparing the image quality of different sharpening intensities, the optimal sharpening intensity is selected to ensure that the edge clarity achieves the best effect without affecting other image features. Finally, the processed edge information data is uniformly processed to generate edge optimization data.

[0111] The texture distribution optimization submodule analyzes the texture details in the area and the texture distribution differences in the surrounding areas based on the edge optimization data, adjusts the contrast and consistency of the texture, and generates texture optimization data;

[0112] First, by comparing the distribution of texture details in the area, analyzing the texture differences between the local area and the surrounding area, using texture analysis technology (such as grayscale co-occurrence matrix) to extract texture features, for areas with rich texture details, adjust the contrast to improve the texture clarity of the local area, then analyze the texture consistency between regions, adjust the texture transition of different regions, ensure the natural connection between high-contrast areas and low-contrast areas, optimize the texture consistency and contrast of different regions through multiple iterations, ensure the texture distribution of the entire image is coordinated, and finally summarize the optimized texture data to generate texture optimization data.

[0113] The regional detail adjustment submodule analyzes the matching degree of edge and texture features in the region based on the texture optimization data, performs synchronous optimization of the edge and texture, and generates optimized iris image data;

[0114] First, the edge and texture features in the area are analyzed synchronously, and the edge detection and texture analysis methods are used to identify the feature matching degree in the area. For areas with low edge and texture detail matching, synchronous optimization is performed to ensure that the edge sharpness and texture detail performance are coordinated. By adjusting the edge brightness and texture contrast, the detail performance of the two is made consistent. Then, the detail features in different areas are uniformly adjusted to ensure that the detail features of the entire image remain coherent and consistent. Finally, the optimized data of all areas are integrated to generate complete optimized iris image data.

[0115] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. High-precision multi-spectral iris image acquisition system, characterized by: The system comprises: The iris region segmentation module obtains the iris image from the collected face information, analyzes the local texture features of the iris, identifies different structures in the iris, divides the image into regions according to the structural and texture features, and generates iris local structure region data; The spectral response analysis module analyzes the spectral characteristics of each local area according to the iris local structure area data, compares the spectral response of the bands, extracts the spectral change curve of the local area in each band, and generates the spectral response characteristic data of the area; The phase difference compensation module analyzes the phase difference trend of the band according to the spectral response characteristic data of the region, adjusts the delay amount of the band, synchronizes the image details, and generates phase-synchronized image data; The spatial entropy enhancement module calculates the entropy value of each band in the image according to the phase-synchronized image data, compares the entropy value distribution, adjusts the brightness and contrast of the selected area, and generates entropy enhanced area image data; The feature enhancement module extracts features from the entropy enhanced region image data, identifies and analyzes local edge and texture features, integrates the extracted features, and generates iris image feature integration data; The image detail optimization module integrates data according to the iris image features, optimizes and analyzes the edge and texture details of the local area, adjusts the details of the local area, and generates optimized iris image data.

2. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The iris local structure area data includes the texture boundary coordinates of the area, the grayscale level distribution of the detail area and the pixel structure information of each area. The spectral response characteristic data of the area includes the spectral intensity under multiple bands, the spectral absorption peak information of the local area and the spectral reflectance ratio between bands. The phase synchronization image data includes the aligned image pixel position information, the phase adjustment coefficient after band synchronization and the synchronization state identification of the area. The entropy enhancement area image data includes the spatial entropy value distribution of the enhanced area, the adjusted contrast level and the detail enhancement parameters of the area. The iris image feature data includes the edge contour coordinates, the detail change rate of the local texture and the edge feature comparison results under multiple spectra. The optimized iris image data includes the optimized edge sharpening degree, texture consistency parameters and regional detail matching score.

3. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The iris region segmentation module comprises: The structural feature extraction submodule identifies the area of ​​structural changes in the iris image and extracts the key structural layers in the image through boundary analysis. The key structural layers are the information of iris texture, iris edge, pupil and iris junction, and generates iris structural feature data. The local segmentation processing submodule divides the iris image into multiple regions based on the iris structural feature data and with reference to the identified structural boundaries, and adjusts the segmentation boundaries by comparing the structural changes within the regions to generate iris local division data; The regional texture analysis submodule analyzes the distribution of texture features in the region based on the iris local segmentation data, identifies the texture change trend between regions, compares the texture difference of each region, and generates iris local structure region data.

4. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The spectral response analysis module comprises: The spectral reflection information acquisition submodule acquires the reflection information of the local area under multiple spectral bands based on the iris local structure area data, analyzes the reflection characteristics of the area under each band, extracts the reflection intensity information of the local area, and generates spectral reflection data; The spectral absorption information analysis submodule compares the reflection intensity of the region under multiple bands based on the spectral reflection data, calculates the spectral absorption value of the local area, analyzes the absorption change trend under the current band, and generates spectral absorption data; The spectral change analysis submodule extracts the spectral change trend between bands based on the spectral absorption data, identifies the spectral characteristics of the region by comparing the spectral response differences in the local area, and generates the spectral response characteristic data of the region.

5. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The phase difference compensation module comprises: The light delay measurement submodule analyzes the light propagation path of the band based on the spectral response characteristic data of the area, measures the delay time of each band of light on the iris surface and each area, compares the light delay changes of the band to extract the delay time difference of each band, and generates the light delay data of the band; The phase difference calculation submodule calculates the phase difference between adjacent bands based on the light delay data of the bands and the delay difference information of the bands, analyzes the change trend of the phase difference between the bands, and generates the band phase difference data; The detail synchronization processing submodule performs detail synchronization processing based on the band phase difference data by adjusting the phase difference value of the iris image in each band to generate phase synchronization image data.

6. The high-precision multi-spectral iris image acquisition system according to claim 5, characterized in that: According to the delay difference information of the band, according to the formula: Calculate the phase difference Δφ between adjacent bands; In the formula, Δt represents the light delay time difference between bands, T represents the period of the light wave, and λ1 and λ2 represent the central wavelengths of two adjacent bands respectively.

7. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The spatial entropy enhancement module includes: The entropy value calculation submodule performs information analysis on the local area of ​​each band iris image based on the phase synchronization image data, calculates the entropy value of the area in the image, identifies the entropy value distribution of each band image, and generates band entropy value distribution data; The high entropy region identification submodule extracts the region with information content based on the band entropy value distribution data by comparing the entropy value distribution, analyzes the boundary and position of the selected entropy value region, and generates the selected entropy value region data; The detail enhancement processing submodule optimizes the detail information of the region by adjusting the brightness and contrast based on the selected entropy value region data to generate entropy enhanced region image data.

8. The high-precision multi-spectral iris image acquisition system according to claim 7, characterized in that: Refer to the local area information of the reference image, according to the formula: Calculate the entropy value H(X) of the local area; In the formula, P(x i ) means the pixel intensity is x i The probability of i being the index of pixel intensity, log2P(x i ) is the logarithm of the intensity probability, m is the total number of pixels to be analyzed, Δx j is the change in the position of pixel j, d j It represents the distance from pixel j to the center of the image, and Z is a normalization constant.

9. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The feature enhancement module comprises: The edge information extraction submodule extracts the edge of the iris image of the entropy enhanced region image data, analyzes the continuity of the edge of the iris image, identifies the edge boundary through the brightness difference of the edge, extracts the starting and ending positions of the edge, and generates edge information data; The texture feature recognition submodule compares the texture changes in the region based on the edge information data, identifies the direction and density of the texture under multiple bands, and generates texture feature data; The multi-band feature fusion submodule integrates the feature information based on the texture feature data to generate iris image feature integration data.

10. The high-precision multi-spectral iris image acquisition system according to claim 1, characterized in that: The image detail optimization module comprises: The edge optimization analysis submodule integrates data based on the iris image features, analyzes the edge contour in the region and compares it with the boundary of the adjacent region, adjusts the edge sharpening strength, and generates edge optimization data; The texture distribution optimization submodule analyzes the difference between the texture details in the region and the texture distribution in the surrounding region based on the edge optimization data, adjusts the contrast and consistency of the texture, and generates texture optimization data; The regional detail adjustment submodule analyzes the matching degree of the edge and texture features in the region based on the texture optimization data, performs synchronous optimization of the edge and texture, and generates optimized iris image data.

Images