Face recognition method and system based on multi-dimensional features and program product

Abstract

The invention belongs to the technical field of face recognition, and particularly discloses a multi-dimensional feature-based face recognition method and system and a program product, and the method comprises the steps: collecting a face image of a target object, carrying out the enhancement processing, obtaining an enhanced face image, carrying out the feature extraction of three dimensions, i.e., gray scale features, texture features and contour features, based on the enhanced face image, and obtaining a face recognition result. And then determining the similarity between the face image of the target object and the corresponding face image sample in the database under the three-dimensional features, and finally judging the face recognition result of the target object based on the weighted summation similarity, so that accurate and efficient face recognition can be realized. According to the invention, through the face recognition process in combination with multi-dimensional feature analysis, the limitation of single feature dimension face recognition can be effectively solved, the accuracy and robustness of face recognition are improved, and the reliability of a corresponding face recognition system is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of face recognition, and particularly relates to a face recognition method, system and program product based on multi-dimensional features. Background Art

[0002] Face recognition is a technology for identity recognition based on facial features. Usually, a camera or webcam is used to collect photos or videos containing human faces. Through face detection and analysis, the human faces in the pictures or videos are recognized and identity verification is performed. Most traditional face recognition technologies analyze and detect using single-dimensional facial image features, which have certain technical limitations. For example, changes in environmental light or expressions can lead to face recognition failures or errors. The accuracy, adaptability, and reliability of this single-feature-dimensional face recognition method still need to be improved. Summary of the Invention

[0003] The purpose of the present invention is to provide a face recognition method, system and program product based on multi-dimensional features to solve the above problems existing in the prior art.

[0004] To achieve the above purpose, the present invention adopts the following technical solutions: In the first aspect, a face recognition method based on multi-dimensional features is provided, including: Obtain the face image of the target object collected by the image acquisition end, and perform geometric transformation processing on the face image to obtain a preprocessed face image; Perform image filtering and image enhancement processing on the preprocessed face image to obtain an enhanced face image; Perform gray-scale transformation processing on the enhanced face image to obtain a gray-scale face image, and extract a gray-scale feature set from the gray-scale face image; Extract texture features from the enhanced face image based on a set of Gabor filters to obtain a texture feature set; Perform edge detection and contour extraction on the enhanced face image to determine a set of facial contours, and extract a contour feature set from the set of facial contours; Traverse each sample face image in the database, as well as the corresponding sample contour feature set, sample gray-scale feature set, and sample texture feature set associated with each sample face image; Calculate the first similarity between the gray-scale feature set and the sample gray-scale feature set associated with each sample face image, the second similarity between the texture feature set and the sample texture feature set associated with each sample face image, and the third similarity between the contour feature set and the sample contour feature set associated with each sample face image; Perform weighted summation on the first similarity, the second similarity, and the third similarity to obtain the overall similarity between the face image of the target object and the corresponding sample face image; When the overall similarity between the face image of the target object and a certain sample face image in the database exceeds the set similarity threshold, it is determined that the face recognition of the target object passes.

[0005] In a possible design, the geometric transformation processing of the face image to obtain the preprocessed face image includes: performing image rotation and image scaling processing on the face image to obtain a preprocessed face image with a fixed resolution.

[0006] In a possible design, the image filtering and image enhancement processing of the preprocessed face image to obtain the enhanced face image includes: Performing image filtering processing on the preprocessed face image using the Gaussian filtering algorithm to obtain the filtered face image; Performing histogram equalization image enhancement processing on the filtered face image to obtain the enhanced face image.

[0007] In a possible design, the gray-scale transformation processing of the enhanced face image to obtain the gray-scale face image and extracting the gray-scale feature set from the gray-scale face image includes: Substituting the brightness values of the RGB three color channels of each pixel point in the enhanced face image into a preset gray-scale conversion formula for calculation to obtain the first gray-scale value of each pixel point. The gray-scale conversion formula is H = 0.30R + 0.59G + 0.11B, where H represents the first gray-scale value, and R, G, and B are the brightness values of the RGB three color channels respectively; Calculating the gray-scale mean value of the enhanced face image based on the first gray-scale value of each pixel point, and subtracting the gray-scale mean value from the first gray-scale value of each pixel point to obtain the gray-scale difference value of each pixel point; Taking the sum of the gray-scale difference values of each row of pixel points in the enhanced face image as the gray-scale feature value of that row of pixel points; Combining the gray-scale feature values of each row of pixel points in the enhanced face image in the order of the corresponding rows to obtain the gray-scale feature set.

[0008] In a possible design, the texture feature extraction of the enhanced face image based on a set Gabor filter bank to obtain the texture feature set includes: Selecting Gabor filters with M scale parameters and N direction parameters to form a Gabor filter bank; Dividing the enhanced face image into L×L image blocks, and respectively importing the L×L image blocks into the Gabor filter bank, so that the Gabor filter bank performs Gabor filtering on each image block respectively to obtain the corresponding M×N texture feature values; Using the M×N texture feature values of each image block to form the texture feature value sequence of the corresponding image block, and using the texture feature value sequences of each image block to form the texture feature set.

[0009] In a possible design, the edge detection and contour extraction of the enhanced face image are performed to determine the face contour set, and the contour feature set is extracted from the face contour set, including: Substitute the brightness values of the RGB three color channels of each pixel point in the enhanced face image into a preset gray-scale processing formula for calculation to obtain the second gray-scale value of each pixel point. The gray-scale processing formula is D = q1R + q2G + q3B, where D represents the second gray-scale value, R, G, and B are the brightness values of the RGB three color channels respectively, and q1, q2, and q3 are the set first weight coefficient, second weight coefficient, and third weight coefficient respectively; Construct a gray-scale conversion image based on the second gray-scale value of each pixel point, and use the Prewitt operator to perform edge detection on the gray-scale conversion image to extract several closed edge contours to form the face contour set; Take the closed edge contour with the largest contour area and within a set first area range in the face contour set as the face contour, take the closed edge contour with the contour area within a set second area range and the aspect ratio of the minimum circumscribed rectangle of the contour within a set first aspect ratio range in the face contour set as the nose contour, take the closed edge contour with the contour area within a set third area range and the aspect ratio of the minimum circumscribed rectangle of the contour within a set second aspect ratio range in the face contour set as the mouth contour, and take the closed edge contour with the contour area within a set fourth area range and the aspect ratio of the minimum circumscribed rectangle of the contour within a set third aspect ratio range in the face contour set as the eye contour; Perform corner detection on the mouth contour to obtain the first corner point of the mouth contour, and perform corner detection on the eye contour to obtain the second corner point of the eye contour; Taking the contour center point of the nose contour as the reference point, construct several first connection lines from the reference point to the edge of the face contour, and the angle between adjacent two first connection lines is fixed. Construct the second connection lines from the reference point to each first corner point, and construct the third connection lines from the reference point to each second corner point; Determine the image distances of each first connection line, second connection line, and third connection line, and arrange and combine the image distances of each first connection line, second connection line, and third connection line in ascending order to obtain the contour feature set.

[0010] In a possible design, the calculation of the first similarity between the gray-scale feature set and the sample gray-scale feature sets associated with each sample face image, the second similarity between the texture feature set and the sample texture feature sets associated with each sample face image, and the third similarity between the contour feature set and the sample contour feature sets associated with each sample face image includes: Calculate the first Euclidean distance between the grayscale feature set and the sample grayscale feature sets associated with each sample face image, and determine the first similarity between the grayscale feature set and the sample grayscale feature sets associated with each sample face image according to the first Euclidean distance; Calculate the second Euclidean distance between the texture feature set and the sample texture feature sets associated with each sample face image, and determine the second similarity between the texture feature set and the sample texture feature sets associated with each sample face image according to the second Euclidean distance; Calculate the third Euclidean distance between the contour feature set and the sample contour feature sets associated with each sample face image, and determine the third similarity between the contour feature set and the sample contour feature sets associated with each sample face image according to the third Euclidean distance.

[0011] In a second aspect, a face recognition system based on multi-dimensional features is provided, including an image acquisition unit, an image enhancement unit, a first extraction unit, a second extraction unit, a third extraction unit, a sample traversal unit, a comparison calculation unit, a weighted summation unit, and a matching recognition unit, where: The image acquisition unit is configured to acquire a face image of a target object collected by an image acquisition terminal, and perform geometric transformation processing on the face image to obtain a preprocessed face image; The image enhancement unit is configured to perform image filtering and image enhancement processing on the preprocessed face image to obtain an enhanced face image; The first extraction unit is configured to perform grayscale transformation processing on the enhanced face image to obtain a grayscale face image, and extract a grayscale feature set from the grayscale face image; The second extraction unit is configured to extract texture features from the enhanced face image based on a set of Gabor filters to obtain a texture feature set; The third extraction unit is configured to perform edge detection and contour extraction on the enhanced face image to determine a set of facial contours, and extract a contour feature set from the set of facial contours; The sample traversal unit is configured to traverse each sample face image in the database, as well as the corresponding sample contour feature set, sample grayscale feature set, and sample texture feature set associated with each sample face image; The comparison calculation unit is configured to calculate the first similarity between the grayscale feature set and the sample grayscale feature sets associated with each sample face image, the second similarity between the texture feature set and the sample texture feature sets associated with each sample face image, and the third similarity between the contour feature set and the sample contour feature sets associated with each sample face image, respectively; The weighted summation unit is configured to perform weighted summation on the first similarity, the second similarity, and the third similarity to obtain the overall similarity between the face image of the target object and the corresponding sample face image; A matching and recognition unit, configured to determine that the face recognition of the target object passes when the overall similarity between the face image of the target object and a certain sample face image in the database exceeds a set similarity threshold.

[0012] In a third aspect, a face recognition system based on multi-dimensional features is provided, including: A memory for storing instructions; A processor for reading the instructions stored in the memory and executing any one of the methods in the first aspect according to the instructions.

[0013] In a fourth aspect, a computer-readable storage medium is provided, on which instructions are stored. When the instructions run on a computer, the computer is made to execute any one of the methods in the first aspect. At the same time, a computer program product is also provided, which, when running on a computer, executes any one of the methods in the first aspect.

[0014] Advantageous effects: By collecting the face image of the target object for enhancement processing to obtain an enhanced face image, then extracting features in three dimensions of grayscale features, texture features, and contour features based on the enhanced face image, and then determining the similarity between the face image of the target object and the corresponding face image sample in the database under the three-dimensional features, and finally determining the face recognition result of the target object based on the weighted sum similarity, accurate and efficient face recognition can be achieved. Through the face recognition process combining multi-dimensional feature analysis, the limitations of traditional single-feature-dimensional face recognition can be effectively solved, the accuracy and robustness of face recognition can be improved, and the reliability of the corresponding face recognition system can be enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 It is a schematic diagram of the steps of the method in Embodiment 1 of the present invention; Figure 2 It is a schematic diagram of the composition of the system in Embodiment 2 of the present invention; Figure 3 It is a schematic diagram of the composition of the system in Embodiment 3 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] It should be noted here that the description of these embodiments is for helping to understand the present invention, but does not constitute a limitation to the present invention. The specific structural and functional details disclosed herein are only used to describe the exemplary embodiments of the present invention. However, the present invention can be embodied in many alternative forms and should not be construed as being limited to the embodiments set forth herein.

[0018] It should be understood that, unless otherwise clearly specified and defined, the corresponding terms should be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be an electrical connection, a direct connection, an indirect connection through an intermediate medium, or a communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments can be understood according to specific situations.

[0019] Specific details are provided in the following description to facilitate a complete understanding of the exemplary embodiments. However, those of ordinary skill in the art should understand that the exemplary embodiments can be implemented without these specific details. For example, the device can be shown in a block diagram to avoid making the example unclear with unnecessary details. In other embodiments, well-known processes, structures, and technologies can be shown without unnecessary details to avoid making the embodiments unclear.

[0020] Embodiment 1: This embodiment provides a face recognition method based on multi-dimensional features, which can be applied to a corresponding face recognition system, such as Figure 1 As shown, the method includes the following steps: S1. Obtain the face image of the target object collected by the image acquisition end, and perform geometric transformation processing on the face image to obtain the preprocessed face image.

[0021] In specific implementation, first collect the face image of the target object through the image acquisition end, and then transmit the face image of the target object to the face recognition system. After the face recognition system obtains the face image of the target object, it can perform corresponding geometric transformation processing on the face image, such as performing image rotation and image scaling processing on the face image to obtain a preprocessed face image with a fixed resolution.

[0022] S2. Perform image filtering and image enhancement processing on the preprocessed face image to obtain an enhanced face image.

[0023] In specific implementation, the system can use the Gaussian filtering algorithm to perform image filtering processing on the preprocessed face image to obtain a filtered face image; then perform histogram equalization image enhancement processing on the filtered face image to obtain an enhanced face image.

[0024] S3. Perform grayscale transformation on the enhanced face image to obtain a grayscale face image, and extract a grayscale feature set from the grayscale face image.

[0025] Specifically, the system can substitute the brightness values of the RGB three-color channels of each pixel point in the enhanced face image into a preset grayscale conversion formula for calculation to obtain the first grayscale value of each pixel point. The grayscale conversion formula is H = 0.30R + 0.59G + 0.11B, where H represents the first grayscale value, and R, G, and B are the brightness values of the RGB three-color channels respectively. Then, calculate the grayscale mean value of the enhanced face image based on the first grayscale value of each pixel point, and subtract the grayscale mean value from the first grayscale value of each pixel point to obtain the grayscale difference value of each pixel point. Then, take the sum of the grayscale difference values of each row of pixel points in the enhanced face image as the grayscale feature value of that row of pixel points, and combine the grayscale feature values of each row of pixel points in the enhanced face image in the order of the corresponding rows (such as from top to bottom) to obtain a grayscale feature set.

[0026] S4. Extract texture features from the enhanced face image based on a set of Gabor filters to obtain a texture feature set.

[0027] Specifically, the system can pre-construct corresponding Gabor filters and select a Gabor filter group composed of M scale parameters and N direction parameters of Gabor filters; then divide the enhanced face image into L×L image blocks, and import the L×L image blocks into the Gabor filter group respectively, so that the Gabor filter group performs Gabor filtering on each image block respectively to obtain the corresponding M×N texture feature values; then use the M×N texture feature values of each image block to form a texture feature value sequence corresponding to the image block, and use the texture feature value sequences of each image block to form a texture feature set.

[0028] S5. Perform edge detection and contour extraction on the enhanced face image to determine a set of facial contours, and extract a contour feature set from the set of facial contours.

[0029] Specifically, the system can first substitute the brightness values of the RGB three-color channels of each pixel point in the enhanced face image into a preset grayscale processing formula for calculation to obtain the second grayscale value of each pixel point. The grayscale processing formula is D = q1R + q2G + q3B, where D represents the second grayscale value, R, G, and B are the brightness values of the RGB three-color channels respectively, and q1, q2, and q3 are the set first weight coefficient, second weight coefficient, and third weight coefficient respectively. Then, construct a grayscale conversion image based on the second grayscale value of each pixel point, and use the Prewitt operator to perform edge detection on the grayscale conversion image to extract several closed edge contours to form a set of facial contours.

[0030] Then, the closed edge contour with the largest contour area in the face contour set and within the set first area range can be used as the face contour. The closed edge contour with the contour area within the set second area range and the aspect ratio of the minimum circumscribed rectangle of the contour within the set first aspect ratio range in the face contour set can be used as the nose contour. The closed edge contour with the contour area within the set third area range and the aspect ratio of the minimum circumscribed rectangle of the contour within the set second aspect ratio range in the face contour set can be used as the mouth contour. The closed edge contour with the contour area within the set fourth area range and the aspect ratio of the minimum circumscribed rectangle of the contour within the set third aspect ratio range in the face contour set can be used as the eye contour.

[0031] Next, perform corner detection on the mouth contour to obtain the first corner points of the mouth contour, and perform corner detection on the eye contour to obtain the second corner points of the eye contour. Taking the contour center point of the nose contour as the reference point, construct a number of first connection lines from the reference point to the edge of the face contour, and the angle between adjacent two first connection lines is fixed. Construct the second connection lines from the reference point to each first corner point, and construct the third connection lines from the reference point to each second corner point.

[0032] Finally, determine the image distances of each first connection line, second connection line, and third connection line, and arrange and combine the image distances of each first connection line, second connection line, and third connection line in ascending order to obtain the contour feature set.

[0033] S6. Traverse each sample face image in the database, as well as the sample contour feature set, sample grayscale feature set, and sample texture feature set associated with each sample face image.

[0034] In specific implementation, the system can traverse each sample face image in the database, as well as the sample contour feature set, sample grayscale feature set, and sample texture feature set associated with each sample face image. A number of sample face images are pre-stored in the database, as well as the sample contour feature set, sample grayscale feature set, and sample texture feature set associated with each sample face image obtained by the same method.

[0035] S7. Calculate the first similarity between the grayscale feature set and the sample grayscale feature set associated with each sample face image, the second similarity between the texture feature set and the sample texture feature set associated with each sample face image, and the third similarity between the contour feature set and the sample contour feature set associated with each sample face image.

[0036] In specific implementation, the system can calculate the first Euclidean distance between the grayscale feature set and the sample grayscale feature sets associated with each sample face image, and determine the first similarity between the grayscale feature set and the sample grayscale feature sets associated with each sample face image according to the first Euclidean distance. It can calculate the second Euclidean distance between the texture feature set and the sample texture feature sets associated with each sample face image, and determine the second similarity between the texture feature set and the sample texture feature sets associated with each sample face image according to the second Euclidean distance. It can calculate the third Euclidean distance between the contour feature set and the sample contour feature sets associated with each sample face image, and determine the third similarity between the contour feature set and the sample contour feature sets associated with each sample face image according to the third Euclidean distance.

[0037] S8. Perform a weighted sum of the first similarity, the second similarity, and the third similarity to obtain the overall similarity between the face image of the target object and the corresponding sample face image.

[0038] In specific implementation, the system can perform a weighted sum of the first similarity, the second similarity, and the third similarity to obtain the overall similarity between the face image of the target object and the corresponding sample face image. Exemplarily, the weighted sum formula can be set as S 总 = 0.4S1 + 0.2S2 + 0.4S3, where S 总 is the overall similarity, S1 is the first similarity, S2 is the second similarity, and S3 is the third similarity.

[0039] S9. When the overall similarity between the face image of the target object and a certain sample face image in the database exceeds the set similarity threshold, it is determined that the face recognition of the target object passes.

[0040] In specific implementation, when the overall similarity between the face image of the target object and a certain sample face image in the database exceeds the set similarity threshold, the system can determine that the face recognition of the target object passes, output information indicating that the face recognition passes, and at the same time, it can also determine the identity information associated with the corresponding sample face image, and synchronously output the identity information corresponding to the matching sample face image as the identity information of the target object.

[0041] The method of this embodiment can effectively solve the limitations of traditional single - feature - dimension face recognition through the face recognition process combining multi - dimensional feature analysis, improve the accuracy and robustness of face recognition, and enhance the reliability of the corresponding face recognition system.

[0042] Embodiment 2: This embodiment provides a face recognition system based on multi - dimensional features, as Figure 2 shown, including an image acquisition unit, an image enhancement unit, a first extraction unit, a second extraction unit, a third extraction unit, a sample traversal unit, a comparison calculation unit, a weighted sum unit, and a matching recognition unit, where: An image acquisition unit, configured to acquire a face image of a target object collected by an image acquisition end, and perform geometric transformation processing on the face image to obtain a preprocessed face image; An image enhancement unit, configured to perform image filtering and image enhancement processing on the preprocessed face image to obtain an enhanced face image; A first extraction unit, configured to perform gray-scale transformation processing on the enhanced face image to obtain a gray-scale face image, and extract a gray-scale feature set from the gray-scale face image; A second extraction unit, configured to perform texture feature extraction on the enhanced face image based on a set Gabor filter bank to obtain a texture feature set; A third extraction unit, configured to perform edge detection and contour extraction on the enhanced face image, determine a set of face contours, and extract a contour feature set from the set of face contours; A sample traversal unit, configured to traverse various sample face images in a database, as well as the corresponding sample contour feature sets, sample gray-scale feature sets, and sample texture feature sets associated with the various sample face images; A comparison calculation unit, configured to calculate a first similarity between the gray-scale feature set and the sample gray-scale feature sets associated with the various sample face images, a second similarity between the texture feature set and the sample texture feature sets associated with the various sample face images, and a third similarity between the contour feature set and the sample contour feature sets associated with the various sample face images; A weighted summation unit, configured to perform weighted summation on the first similarity, the second similarity, and the third similarity to obtain an overall similarity between the face image of the target object and the corresponding sample face image; A matching and recognition unit, configured to determine that the face recognition of the target object passes when the overall similarity between the face image of the target object and a certain sample face image in the database exceeds a set similarity threshold.

[0043] Embodiment 3: This embodiment provides a face recognition system based on multi-dimensional features. As Figure 3 shown, at the hardware level, it includes: A data interface, configured to establish data docking between a processor and an image acquisition end; A memory, configured to store instructions; A processor, configured to read the instructions stored in the memory and execute the multi-dimensional feature-based face recognition method in Embodiment 1 according to the instructions.

[0044] Optionally, the system further includes an internal bus through which the processor can be interconnected with the memory and the data interface. The internal bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0045] The memory can include, but is not limited to, a random access memory (RAM), a read only memory (ROM), a flash memory, a first input first output (FIFO) memory, and / or a first in last out (FILO) memory, etc. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0046] Embodiment 4: This embodiment provides a computer-readable storage medium with instructions stored thereon. When the instructions are run on a computer, the computer is caused to execute the multi-dimensional feature-based face recognition method in Embodiment 1. Among them, the computer-readable storage medium refers to a carrier for storing data, and can include, but is not limited to, a floppy disk, an optical disc, a hard disk, a flash memory, a USB flash drive, and / or a memory stick, etc. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0047] This embodiment also provides a computer program product. When the computer program product runs on a computer, it executes the face recognition method based on multi-dimensional features in Embodiment 1. Among them, the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0048] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A face recognition method based on multi-dimensional features, characterized in that: include: Acquire the face image of the target object collected by the image acquisition end, and perform geometric transformation processing on the face image to obtain a preprocessed face image; Performing image filtering and image enhancement processing on the preprocessed face image to obtain an enhanced face image; Performing grayscale transformation processing on the enhanced face image to obtain a grayscale face image, and extracting a grayscale feature set from the grayscale face image; Based on the set Gabor filter group, texture features of the enhanced face image are extracted to obtain a texture feature set; Perform edge detection and contour extraction on the enhanced face image, determine a face contour set, and extract a contour feature set from the face contour set; Traversing each sample face image in the database, and the sample contour feature set, sample grayscale feature set and sample texture feature set corresponding to each sample face image; Calculating respectively a first similarity between the grayscale feature set and a sample grayscale feature set associated with each sample face image, a second similarity between the texture feature set and a sample texture feature set associated with each sample face image, and a third similarity between the contour feature set and a sample contour feature set associated with each sample face image; Performing weighted summation on the first similarity, the second similarity, and the third similarity to obtain an overall similarity between the face image of the target object and the corresponding sample face image; When the overall similarity between the face image of the target object and a sample face image in the database exceeds the set similarity threshold, the face recognition of the target object is determined to be passed.

2. The face recognition method based on multi-dimensional features according to claim 1, characterized in that: The method of performing geometric transformation processing on the face image to obtain the preprocessed face image includes: performing image rotation and image scaling processing on the face image to obtain the preprocessed face image with a fixed resolution.

3. The face recognition method based on multi-dimensional features according to claim 1, characterized in that: The step of performing image filtering and image enhancement processing on the pre-processed face image to obtain an enhanced face image includes: Using Gaussian filtering algorithm to perform image filtering on the preprocessed face image to obtain a filtered face image; The filtered face image is subjected to histogram equalization image enhancement processing to obtain an enhanced face image.

4. The face recognition method based on multi-dimensional features according to claim 1, characterized in that: The grayscale transformation process is performed on the enhanced face image to obtain a grayscale face image, and a grayscale feature set is extracted from the grayscale face image, including: Substituting the RGB three-color channel brightness value of each pixel in the enhanced face image into a preset grayscale conversion formula for calculation, to obtain the first grayscale value of each pixel, the grayscale conversion formula is H=0.30R+0.59G+0.11B, where H represents the first grayscale value, and R, G and B are the RGB three-color channel brightness values ​​respectively; Calculating a grayscale mean value of the enhanced face image based on the first grayscale value of each pixel, and subtracting the grayscale mean value from the first grayscale value of each pixel to obtain a grayscale difference value of each pixel; The sum of the grayscale differences of each row of pixels in the enhanced face image is used as the grayscale feature value of the pixel in that row; The grayscale feature values ​​of the pixels in each row of the enhanced face image are combined in the order of the corresponding rows to obtain a grayscale feature set.

5. The face recognition method based on multi-dimensional features according to claim 1, characterized in that: The texture feature extraction of the enhanced face image based on the set Gabor filter group is performed to obtain a texture feature set, including: Select Gabor filters with M scale parameters and N direction parameters to form a Gabor filter bank; The enhanced face image is divided into L×L image blocks, and the L×L image blocks are respectively introduced into the Gabor filter group, so that the Gabor filter group performs Gabor filtering on each image block respectively to obtain corresponding M×N texture feature values; The M×N texture feature values ​​of each image block are used to form a texture feature value sequence of the corresponding image block, and the texture feature value sequence of each image block is used to form a texture feature set.

6. The face recognition method based on multi-dimensional features according to claim 1, characterized in that: The method of performing edge detection and contour extraction on the enhanced face image, determining a face contour set, and extracting a contour feature set from the face contour set includes: Substituting the RGB three-color channel brightness value of each pixel in the enhanced face image into a preset grayscale processing formula for calculation, obtaining a second grayscale value of each pixel, the grayscale processing formula is D=q1R+q2G+q3B, wherein D represents the second grayscale value, R, G and B are the RGB three-color channel brightness values, and q1, q2 and q3 are the set first weight coefficient, second weight coefficient and third weight coefficient respectively; A grayscale conversion image is constructed based on the second grayscale value of each pixel, and the Prewitt operator is used to perform edge detection on the grayscale conversion image to extract a number of closed edge contours to form a facial contour set; The closed edge contour with the largest contour area in the facial contour set and within the set first area range is taken as the face contour, the closed edge contour with the contour area in the facial contour set within the set second area range and the aspect ratio of the minimum circumscribed rectangle of the contour within the set first aspect ratio range is taken as the nose contour, the closed edge contour with the contour area in the facial contour set within the set third area range and the aspect ratio of the minimum circumscribed rectangle of the contour within the set second aspect ratio range is taken as the mouth contour, and the closed edge contour with the contour area in the facial contour set within the set fourth area range and the aspect ratio of the minimum circumscribed rectangle of the contour within the set third aspect ratio range is taken as the eye contour; Perform corner point detection on the mouth contour to obtain the first corner point of the mouth contour, and perform corner point detection on the eye contour to obtain the second corner point of the eye contour; Taking the center point of the nose contour as the reference point, constructing several first connecting lines from the reference point to the edge of the face contour, with the angle between two adjacent first connecting lines being fixed, constructing second connecting lines from the reference point to each first corner point, and constructing third connecting lines from the reference point to each second corner point; The image distances of the first connecting lines, the second connecting lines and the third connecting lines are determined, and the image distances of the first connecting lines, the second connecting lines and the third connecting lines are arranged and combined in order from small to large to obtain a contour feature set.

7. The face recognition method based on multi-dimensional features according to claim 1, characterized in that: The respectively calculating a first similarity between the grayscale feature set and a sample grayscale feature set associated with each sample face image, a second similarity between the texture feature set and a sample texture feature set associated with each sample face image, and a third similarity between the contour feature set and a sample contour feature set associated with each sample face image, comprises: Calculating a first Euclidean distance between the grayscale feature set and a sample grayscale feature set associated with each sample face image, and determining a first similarity between the grayscale feature set and the sample grayscale feature set associated with each sample face image according to the first Euclidean distance; Calculating a second Euclidean distance between the texture feature set and a sample texture feature set associated with each sample face image, and determining a second similarity between the texture feature set and the sample texture feature set associated with each sample face image according to the second Euclidean distance; The third Euclidean distance between the contour feature set and the sample contour feature set associated with each sample face image is calculated, and the third similarity between the contour feature set and the sample contour feature set associated with each sample face image is determined according to the third Euclidean distance.

8. A face recognition system based on multi-dimensional features, characterized in that: It includes an image acquisition unit, an image enhancement unit, a first extraction unit, a second extraction unit, a third extraction unit, a sample traversal unit, a comparison calculation unit, a weighted summation unit and a matching recognition unit, wherein: An image acquisition unit is used to acquire a face image of a target object acquired by an image acquisition terminal, and to perform geometric transformation processing on the face image to obtain a preprocessed face image; An image enhancement unit, used for performing image filtering and image enhancement processing on the pre-processed face image to obtain an enhanced face image; A first extraction unit is used to perform grayscale transformation processing on the enhanced face image to obtain a grayscale face image, and extract a grayscale feature set from the grayscale face image; The second extraction unit is used to extract texture features of the enhanced face image based on a set Gabor filter group to obtain a texture feature set; A third extraction unit is used to perform edge detection and contour extraction on the enhanced face image, determine a face contour set, and extract a contour feature set from the face contour set; A sample traversal unit, used for traversing each sample face image in the database, and a sample contour feature set, a sample grayscale feature set and a sample texture feature set associated with each sample face image; a comparison calculation unit, used to respectively calculate a first similarity between the grayscale feature set and a sample grayscale feature set associated with each sample face image, a second similarity between the texture feature set and a sample texture feature set associated with each sample face image, and a third similarity between the contour feature set and a sample contour feature set associated with each sample face image; A weighted summing unit, used for performing weighted summing of the first similarity, the second similarity and the third similarity to obtain the overall similarity between the face image of the target object and the corresponding sample face image; The matching recognition unit is used to determine that the face recognition of the target object has passed when the overall similarity between the face image of the target object and a sample face image in the database exceeds a set similarity threshold.

9. A face recognition system based on multi-dimensional features, characterized in that: include: A memory for storing instructions; A processor is used to read the instructions stored in the memory and execute the face recognition method based on multidimensional features according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product runs on a computer, the face recognition method based on multi-dimensional features described in any one of claims 1 to 7 is executed.

Images