A method and system for rapid labeling and extraction of connected components in palm vein recognition

Abstract

This invention discloses a method and system for rapid labeling and extraction of connected components in palm vein recognition, belonging to the field of biometric recognition technology. It includes the following steps: S1. Traverse the pixels of the palm vein binary image, detect the first connected component starting point, and set the pixel value of the starting point as a marker; S2. Stop traversing, and perform an eight-neighbor clockwise search starting from the connected component starting point to confirm the outer boundary of the connected component. Set the pixel value of the outer boundary point as a marker. When the search returns to the starting point, the connected component search ends; S3. Continue traversing the pixels of the palm vein binary image, determining whether a pixel is the starting point of another connected component, an internal point of a connected component, or an internal hole boundary point, and performing corresponding operations; S4. Repeat step S3 until all pixels of the palm vein binary image have been traversed, and the largest connected component is extracted. This invention fills each connected component, eliminating the need for recursion and improving operating efficiency.

Description

Technical Field

[0001] This invention belongs to the field of biometric identification technology, and in particular relates to a method and system for rapid labeling and extraction of connected regions in palm vein identification. Background Technology

[0002] Biometric technology is a technology that automatically identifies individuals based on their own biological characteristics. Among them, palm vein recognition is a common type of biometric technology. It is a highly secure identification technology that identifies individuals by extracting the characteristics of the veins in the palm of their hand.

[0003] In the existing process of palm vein recognition, the eight-neighbor connectivity labeling algorithm is generally used to obtain the connected region with the largest area in the vein image. For example, the palm vein feature extraction and matching method based on eight-neighbor and quadratic matching disclosed in the authorization announcement number CN105426821B includes the following steps: (1) acquiring palm vein images; (2) preprocessing the acquired vein images: binarizing to extract the palm, locating key points of the palm, and normalizing and then using histogram equalization to enhance the vein image; (3) extracting the vein contour using the Niblack threshold segmentation method on the enhanced vein image; (4) extracting feature points using the eight-neighbor algorithm.

[0004] The eight-neighbor connectivity labeling algorithm used in traditional palm vein recognition is inefficient and prone to lag due to the need for recursion. Summary of the Invention

[0005] This invention provides a method and system for fast labeling and extraction of connected components in palm vein recognition, to solve the problems of low efficiency and lag when using the eight-neighbor connected component labeling algorithm to obtain the largest area of ​​connected components in vein images in the existing vein recognition process.

[0006] To solve the above-mentioned technical problems, the technical solution provided by the present invention is as follows:

[0007] This invention relates to a method for rapid labeling and extraction of connected regions in palm vein recognition, comprising the following steps:

[0008] S1. Traverse the pixels of the palm vein binary image, detect the first connected component start point, where the pixel value of the start point is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1, and set the pixel value of the start point as a marker. ;

[0009] S2. Stop traversing and start from the starting point of the connected component, performing a clockwise search of its eight neighbors to identify the outer boundary of the connected component. If the pixel value of a certain pixel is equal to 1, then the pixel is determined to be the outer boundary point of the connected component, and the pixel value of the outer boundary point is set as a marker. When the search returns to the starting point, the search for the connected component ends.

[0010] S3. Continue iterating through the pixels of the palm vein binary image, checking whether each pixel satisfies the following conditions and performing the corresponding operations:

[0011] (1) If the pixel is the starting point of a connected component, then mark the number. Increment the value by 1, and repeat step S2;

[0012] (2) If the pixel is an internal point of the connected domain enclosed by the outer boundary points of the connected domain, then set its pixel value to the label number of the corresponding outer boundary point of the connected domain.

[0013] S4. Repeat step S3 until all pixels in the palm vein binary image have been traversed. Count the number of each marker number, select the connected component with the most occurrences of the same marker number as the maximum connected component, extract the maximum connected component, set the pixel value of each pixel in the maximum connected component to 1 and use it as the palm vein recognition region for palm vein recognition.

[0014] Preferably, the marker number in step S1 The initial value is 2.

[0015] Preferably, in step S2, when the coordinates of the searched boundary point are the same as the coordinates of the starting point, it is determined that the search has returned to the starting point.

[0016] Preferably, in step S3, the formula for determining whether a pixel is an interior point of a connected component and setting its pixel value as the marker of its corresponding exterior boundary point of the connected component is as follows:

[0017]

[0018] In the formula, The row coordinates are The vertical axis is The pixel value of the pixel. The row coordinates are The vertical axis is The pixel value of the pixel. The label number of the pixel within the connected component.

[0019] Preferably, the palm vein image needs to be binarized before step S1.

[0020] Preferably, in step S1, when traversing all pixels of the palm vein binary image, the traversal is performed in the order from left to right and from top to bottom.

[0021] This invention also relates to a system for rapid labeling and extraction of connected components in palm vein recognition, comprising:

[0022] The traversal module is used to traverse the pixels of the palm vein binary image. When the first connected component start point is detected, the pixel value of the start point is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1. The pixel value of the start point is then set as a marker. ;

[0023] The outer boundary point search module performs a clockwise search of the eight neighbors starting from the starting point of the connected component to identify the outer boundary of the connected component. If the pixel value of a certain pixel is equal to 1, the pixel is determined to be an outer boundary point of the connected component, and the pixel value of the outer boundary point is set as a marker. When the search returns to the starting point, the search for the connected component ends.

[0024] The boundary point determination module is used to continue traversing the pixels of the palm vein binary image, determine whether the pixels meet the following conditions, and perform corresponding operations:

[0025] (1) If the pixel is the starting point of a connected component, then mark the number. Increment the value by 1, and repeat step S2;

[0026] (2) If the pixel is an internal point of the connected domain enclosed by the outer boundary points of the connected domain, then set its pixel value to the label number of the corresponding outer boundary point of the connected domain.

[0027] The extraction module is used to count the number of each marker, select the connected component with the most occurrences of the same marker as the maximum connected component, and extract the maximum connected component as the palm vein recognition region for palm vein recognition.

[0028] Compared with the prior art, the technical solution provided by this invention has the following advantages:

[0029] 1. The fast labeling and extraction method for connected components in palm vein recognition involved in this invention detects the starting point of connected components by traversing the pixels in the binary image of the palm vein. The pixel value of the starting point of the connected component is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1. Then, the eight neighboring pixels are searched clockwise from the starting point of the connected component. In this way, the palm vein image can be divided into several connected components without recursion, which improves the running efficiency.

[0030] 2. The fast labeling and extraction method for connected components in palm vein recognition of the present invention selects a pixel with a pixel value equal to 1 and whose eight neighboring pixel values ​​are all less than or equal to 1 as the starting point, and searches for connected components through the starting point. The starting point needs to simultaneously satisfy that the pixel values ​​of the eight neighboring pixel values ​​are all less than or equal to 1, which can avoid using points inside the palm as starting points for boundary labeling.

[0031] 3. The fast labeling and extraction method for connected components in palm vein recognition involved in this invention, when searching for boundary points of connected components, judges the searched pixels. If the pixel is an internal point of the connected component enclosed by the outer boundary points of the connected component, its pixel value is set as the label number of the corresponding outer boundary point of the connected component. If the pixel is an internal hole boundary point, the internal hole boundary point is labeled, and the label number is the largest pixel value among its eight neighbors. In this way, it can avoid labeling the internal holes between fingers as a separate connected component, thereby reducing the cardinality of subsequent connected component statistics and improving the running speed. Attached Figure Description

[0032] Figure 1 This is a block diagram of a system for rapid labeling and extraction of connected components in palm vein recognition. Detailed Implementation

[0033] To further understand the content of this invention, the invention will be described in detail with reference to the embodiments. The following embodiments are used to illustrate the invention, but are not intended to limit the scope of the invention.

[0034] Example 1: This example uses palm vein recognition as an example to describe in detail a method for rapid labeling and extraction of connected components in palm vein recognition, which includes the following steps:

[0035] S0. Binarize the palm vein image to obtain a palm vein binary map, which in this embodiment contains two connected components;

[0036] S1. Traverse the pixels of the palm vein binary image, detect the first connected component start point, the pixel value of the connected component start point is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1, and set the pixel value of the start point as a marker. The specific steps are as follows:

[0037] Traverse the pixels of the palm vein binary image from left to right and top to bottom. When the first pixel with a value of 1 and all eight neighboring pixels with values ​​less than or equal to 1 are detected, this pixel is determined as the starting point of the connected component and is used as the starting point of the connected component. Starting point The coordinates are , the starting point The pixel value is set as the marker number. , tag number The initial value is 2.

[0038] S2. Stop traversing and start from the starting point of the connected component, performing a clockwise search of its eight neighbors to identify the outer boundary of the connected component. If the pixel value of a certain pixel is equal to 1, then the pixel is determined to be the outer boundary point of the connected component, and the pixel value of the outer boundary point is set as a marker. When the search returns to the starting point, the search for the connected component ends. The specific steps are as follows:

[0039] After detecting the starting point of the connected component, stop the traversal and start a clockwise search of the eight neighbors from the starting point to confirm the outer boundary of the connected component. If a certain pixel... If a pixel value is equal to 1, then that pixel is determined to be an outer boundary point of the connected component, and the pixel value of that outer boundary point is set as a marker. ,Right now:

[0040]

[0041] In the formula, and These represent the search results for the first time. The x and y coordinates of the outer boundary points No. Pixel values ​​of the outer boundary points;

[0042] When the search returns to the starting point, that is, when the x and y coordinates of the searched outer boundary point satisfy the formula (3), it indicates that the search of the connected domain has ended;

[0043] .

[0044] Step S2 yields a palm vein image after marking the outer boundary of a connected domain.

[0045] S3. Continue iterating through the pixels of the palm vein binary image, checking whether each pixel satisfies the following conditions and performing the corresponding operations:

[0046] (1) If the pixel is the starting point of a connected component, then mark the number. Increment the value by 1, and repeat step S2;

[0047] (2) If the pixel is an internal point of the connected domain enclosed by the outer boundary points of the connected domain, then set its pixel value to the label number of the corresponding outer boundary point of the connected domain.

[0048] This avoids marking the internal holes between fingers as a separate connected component, thereby reducing the cardinality of subsequent connected component statistics and improving the running speed.

[0049] In the above process, the formula for determining whether a pixel is an interior point of a connected component and setting its pixel value to the label of its corresponding exterior boundary point of the connected component is as follows:

[0050]

[0051] In the formula, The row coordinates are i The vertical axis is jThe pixel value of the pixel. The row coordinates are i The vertical axis is j+ The pixel value of a pixel with a value of 1. The label number of the pixel within the connected component.

[0052] Step S3 yields a palm vein image after marking the outer boundaries of the two connected domains, marking the points inside the connected domains, and marking the boundary points of the internal holes.

[0053] S4. Repeat step S3 until all pixels in the palm vein binary image have been traversed. Count the number of each marker number, select the connected component with the most occurrences of the same marker number as the maximum connected component, extract the maximum connected component, set the pixel value of each pixel in the maximum connected component to 1, and use it as the palm vein recognition region for palm vein recognition. The specific steps are as follows:

[0054] Repeat step S3 until all pixels in the palm vein binary image have been traversed. Count the number of each marker and select the connected component with the most occurrences of the same marker as the largest connected component. Extract the largest connected component, set the pixel value of each pixel in the largest connected component to 1 according to formula (4) and use it as the palm vein recognition area. Use the palm vein recognition area after setting the pixel value of the largest connected component to 1 to perform palm vein recognition.

[0055]

[0056] In the formula, It represents the pixel value of the largest connected component.

[0057] Example 2: Refer to Appendix Figure 1 As shown, this embodiment relates to a system for rapid labeling and extraction of connected components in palm vein recognition, which includes:

[0058] The traversal module is used to traverse the pixels of the palm vein binary image. When the first connected component start point is detected, the pixel value of the start point is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1. The pixel value of the start point is then set as a marker. The traversal module is used to implement the function of step S1 in Example 1.

[0059] The outer boundary point search module performs a clockwise search of the eight neighbors starting from the starting point of the connected component to identify the outer boundary of the connected component. If the pixel value of a certain pixel is equal to 1, the pixel is determined to be an outer boundary point of the connected component, and the pixel value of the outer boundary point is set as a marker. When the search returns to the starting point, the search for the connected component ends. The outer boundary point search module is used to implement the function of step S2 in embodiment 1.

[0060] The boundary point determination module is used to continue traversing the pixels of the palm vein binary image, determine whether the pixel meets the following conditions and perform corresponding operations: (1) If the pixel is the starting point of the connected component, then the marker number is changed. Add 1 to the value and repeat step S2; (2) If the pixel is an internal point of the connected domain surrounded by the outer boundary points of the connected domain, then set its pixel value to the label number of the corresponding outer boundary point of the connected domain.

[0061] The extraction module is used to count the number of each marker, select the connected component with the most occurrences of the same marker as the maximum connected component, and extract the maximum connected component as the palm vein recognition region for palm vein recognition. The extraction module is used to implement the function of step S4 in embodiment 1.

[0062] The present invention has been described in detail above with reference to the embodiments, but the content described is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. All equivalent changes and improvements made in accordance with the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. A method for rapid labeling and extraction of connected components in palm vein recognition, characterized in that: It includes the following steps: S1. Traverse the pixels of the palm vein binary image, detect the first connected component start point, where the pixel value of the start point is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1, and set the pixel value of the start point as a marker. The tag number The initial value is 2; S2. Stop traversing and start from the starting point of the connected component, performing a clockwise search of its eight neighbors to identify the outer boundary of the connected component. If the pixel value of a certain pixel is equal to 1, then the pixel is determined to be the outer boundary point of the connected component, and the pixel value of the outer boundary point is set as a marker. When the search returns to the starting point, the search for the connected component ends. Specifically, when the coordinates of the searched boundary point are the same as the coordinates of the starting point, it is determined that the search has returned to the starting point. S3. Continue iterating through the pixels of the palm vein binary image, checking whether each pixel satisfies the following conditions and performing the corresponding operations: (1) If the pixel is the starting point of a connected component, then mark the number. Increment the value by 1, and repeat step S2; (2) If the pixel is an interior point of the connected region bounded by the outer boundary points of the connected region, then its pixel value is set as the label number of the corresponding outer boundary point of the connected region. The formula for determining whether a pixel is an interior point of the connected region and setting its pixel value as the label number of the corresponding outer boundary point of the connected region is as follows: ； In the formula, The row coordinates are i The vertical axis is j The pixel value of the pixel, Let the row coordinate be i and the y-coordinate be i. j The pixel value of the +1 pixel. The label number of the pixel within the connected component; (3) If the pixel is an internal hole boundary point, then mark the internal hole boundary point. The mark number is the largest pixel value in its eight neighborhoods. The internal hole boundary point is the boundary point of the internal hole in the connected domain. Its pixel value is equal to 1 and there is a neighborhood with a pixel value greater than 1 in its eight neighborhoods. S4. Repeat step S3 until all pixels in the palm vein binary image have been traversed. Count the number of each marker number, select the connected component with the most occurrences of the same marker number as the maximum connected component, extract the maximum connected component, set the pixel value of each pixel in the maximum connected component to 1 and use it as the palm vein recognition region for palm vein recognition.

2. The method for rapid labeling and extraction of connected components in palm vein recognition according to claim 1, characterized in that: Before step S1, the palm vein image needs to be binarized.

3. The method for rapid labeling and extraction of connected components in palm vein recognition according to claim 1, characterized in that: In step S1, when traversing all pixels of the palm vein binary image, the traversal is performed in the order from left to right and from top to bottom.

4. A system for implementing the rapid labeling and extraction method of connected components in palm vein recognition as described in claim 1, characterized in that: It includes: The traversal module is used to traverse the pixels of the palm vein binary image. When the first connected component start point is detected, the pixel value of the start point is equal to 1 and the pixel values ​​of its eight neighboring pixels are all less than or equal to 1. The pixel value of the start point is then set as a marker. The tag number The initial value is 2; The outer boundary point search module performs a clockwise search of the eight neighbors starting from the starting point of the connected component to identify the outer boundary of the connected component. If the pixel value of a certain pixel is equal to 1, the pixel is determined to be an outer boundary point of the connected component, and the pixel value of the outer boundary point is set as a marker. When the search returns to the starting point, the search for the connected component ends. Specifically, when the coordinates of the searched boundary point are the same as the coordinates of the starting point, it is determined that the search has returned to the starting point. The boundary point determination module is used to continue traversing the pixels of the palm vein binary image, determine whether the pixels meet the following conditions, and perform corresponding operations: (1) If the pixel is the starting point of a connected component, then mark the number. Increment the value by 1, and repeat step S2; (2) If the pixel is an interior point of the connected region bounded by the outer boundary points of the connected region, then its pixel value is set as the label number of the corresponding outer boundary point of the connected region. The formula for determining whether a pixel is an interior point of the connected region and setting its pixel value as the label number of the corresponding outer boundary point of the connected region is as follows: ； In the formula, The row coordinates are i The vertical axis is j The pixel value of the pixel, Let the row coordinate be i and the y-coordinate be i. j The pixel value of the +1 pixel. The label number of the pixel within the connected component; (3) If the pixel is an internal hole boundary point, then mark the internal hole boundary point. The mark number is the largest pixel value in its eight neighborhoods. The internal hole boundary point is the boundary point of the internal hole in the connected domain. Its pixel value is equal to 1 and there is a neighborhood with a pixel value greater than 1 in its eight neighborhoods. The extraction module is used to count the number of each marker, select the connected component with the most occurrences of the same marker as the maximum connected component, and extract the maximum connected component as the palm vein recognition region for palm vein recognition.

Images