A method and apparatus for extracting light spots from lattice structured light

Abstract

This invention discloses a method and apparatus for extracting light spots from lattice structured light, relating to the fields of three-dimensional measurement and image processing. The method includes the following steps: obtaining a lattice structured light spot map using a lattice structured light spot projection device; determining the zero-order diffraction center point of the lattice structured light; performing annular segmentation of the lattice structured light map by setting radius parameters; extracting pixel values ​​from different channels in different regions; performing median filtering for noise reduction by setting different thresholds for different regions; superimposing the four obtained grayscale images to obtain the grayscale image of the final image; applying an edge detection algorithm to the grayscale image to determine the edge of each light spot; constructing connected components and displaying the center of the light spot. This invention removes stray light generated by paraxial diffraction, completely extracts the distribution characteristics of the light spot near the zero-order diffraction center point, and preserves the distribution characteristics of the light spot at the edge points to the maximum extent, thus improving the completeness of contour extraction and the accuracy of the light spot center point.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional measurement and image processing, and in particular to a method and apparatus for extracting light spots from lattice structured light. Background Technology

[0002] 3D reconstruction has important applications in fields such as industrial measurement, medical diagnosis, virtual reality, autonomous driving, and defense technology. Optical 3D measurement methods provide crucial front-end data for 3D reconstruction. Currently, mainstream optical 3D measurement methods include projection structured light, time-of-flight methods, and binocular vision. Among these, measurement using lattice structured light obtained through optical diffraction offers advantages over time-of-flight methods and binocular vision, including higher measurement accuracy, stronger anti-interference capabilities, and simpler structure.

[0003] Image processing techniques for spot extraction are an indispensable step in 3D reconstruction. Traditional spot extraction methods include thresholding, edge detection, and morphological processing, which generally perform well on images with clearly defined feature points. However, after close-range structured light illumination, the captured spot image may be overexposed due to excessive light intensity at the image center. Overexposure results in a large area of ​​high-brightness pixels in the image, directly causing edge blurring, and the presence of other noise points further complicates the process.

[0004] The above-mentioned spot extraction methods are ineffective in processing overexposed images, and the noise removal is incomplete or the contour extraction is inaccurate, which directly leads to low accuracy in spot center positioning and may even result in incorrect extraction points. Summary of the Invention

[0005] The purpose of this invention is to provide a method and apparatus for extracting light spots using structured light, thereby solving the problems of inaccurate contour extraction and low center positioning accuracy of light spots caused by overexposure in related technologies. Considering that structured light uses monochromatic light, and that overexposure can lead to the loss of brightness details in parts of the image, overexposed pixels in the image will contain the strongest brightness of the three colors R, G, and B, while other normal pixels will only contain one color information.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A method for extracting light spots from lattice structured light includes: determining the zero-order center point of the collected lattice structured light based on the intensity distribution characteristics of Fraunhofer diffraction; determining three successively enlarged radius values ​​R1, R2, and R3 based on high-brightness pixels caused by overexposure and low-brightness pixels corresponding to high-order diffraction spots, thereby dividing the image into four regions, designated S1, S2, S3, and S4 from the zero-order diffraction center outwards; selecting the B-channel pixel values ​​of the image as the grayscale image of region S1, the G-channel pixel values ​​of the image as the grayscale image of region S2, and the R-channel pixel values ​​of the image as the grayscale images of regions S3 and S4; applying median filtering to region S3 to remove noise by setting a high threshold, and applying median filtering to region S4 to remove noise by setting a low threshold; superimposing the four grayscale images to obtain the grayscale image of the image; applying an edge detection algorithm to the grayscale image to determine the edge of each light spot; constructing connected components and displaying the center of the light spot.

[0008] In the method for extracting light spots from lattice structured light according to the present invention, the Fraunhofer diffraction intensity distribution is as follows:

[0009] ,

[0010] Where U(x,y) represents the light intensity at coordinate point (x,y), k represents the wave vector, λ represents the center wavelength of the laser, z represents the light propagation length, and u(x',y') is the transmission function of the diffraction element used, which needs to be determined according to the diffraction element used in the measurement.

[0011] In the method for extracting light spots using lattice structured light according to the present invention, the radius value R is calculated as follows:

[0012] ,

[0013] Where (x0, y0) are the coordinates of the zeroth order diffraction center point, and (x, y) are the coordinates of the target point.

[0014] It should be specifically noted that in the method for extracting light spots for lattice structured light according to the present invention, different R, G, and B channel pixel values ​​are selected for different regions, taking a 632.8nm red laser as an example. In the measurement of lasers of different color wavelengths, the principle followed in selecting channels in different regions is as follows: region S1, which includes the diffraction zero-order center, selects one of the RGB channels with a significant difference in frequency from the laser light; region S2, adjacent to S1, selects one of the RGB channels with the next smallest difference in frequency from the laser light; and regions S3 and S4 select RGB channels of the same color as the laser. This achieves the optimal selection that balances the quality and noise of paraxial and off-axis light spots.

[0015] In the method for extracting light spots using structured light in this invention, the high threshold and low threshold median filtering threshold values ​​are determined according to the characteristics of the image pixel size, so as to achieve better near-point and far-point denoising effects.

[0016] In a method for extracting light spots using structured light according to the present invention, the edge detection includes: defining two edge templates for detecting edges in directions of +45° and -45° as convolution kernels to perform convolution operations on the image, thereby obtaining the edges of the light spots.

[0017] The present invention also provides a light spot projection device for lattice structured light, comprising: a laser for generating a monochromatic light wave with high brightness and good directionality; an axial cone array, which diffracts the laser beam after irradiation to obtain periodic lattice structured light; a camera for capturing the light spot of the lattice structured light; and a screen for unifying the measurement background and receiving the diffracted light spot of the lattice structured light.

[0018] The beneficial effects of this invention are as follows: By utilizing a region-based, multi-channel discrimination algorithm, it solves the problem of indistinct light spots caused by overexposure near the zero-order diffraction center when the camera captures structured light spots. It removes stray light generated by paraxial diffraction, fully extracts the distribution characteristics of the light spots near the zero-order diffraction center, effectively removes noise from the entire image, and preserves the distribution characteristics of the light spots at the edge points to the maximum extent. While ensuring accuracy, it minimizes data loss, greatly improving the completeness of contour extraction and the accuracy of the light spot center point. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of a light spot extraction device for lattice structured light according to the present invention;

[0021] Figure 2 This is a schematic flowchart of a method for extracting light spots from lattice structured light according to an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram of the low threshold processing result under the conventional binarization algorithm in an embodiment of the present invention;

[0023] Figure 4 This is a schematic diagram of the high threshold processing result under the conventional binarization algorithm in an embodiment of the present invention;

[0024] Figure 5 This is a schematic diagram of the region division after annular multi-channel processing according to an embodiment of the present invention;

[0025] Figure 6 This is a complete schematic diagram of obtaining a spot grayscale image using a ring-shaped multi-channel algorithm according to an embodiment of the present invention;

[0026] Figure 7 This is a schematic diagram of the light spot edge contour detection according to an embodiment of the present invention;

[0027] Figure 8 This is a schematic diagram of the extraction spot center position according to an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] The following is combined with Figures 1 to 8 This invention describes a method and apparatus for extracting light spots from lattice structured light, and includes a comparison of the results obtained by a conventional processing algorithm with the present invention.

[0030] (i) Obtaining the lattice structured light spot image using a lattice structured light spot extraction device

[0031] The experimental optical path is arranged according to Figure 1 Once the setup is complete, turn on the red laser with a center wavelength of 632.8nm. After stabilization, use a camera to expose and capture the dot matrix structured light on the screen to obtain a dot matrix structured light spot image with a resolution of 3456*5184.

[0032] It should be noted that a red laser is used in this step, and theoretically, the spot pattern should only contain R channel information. However, due to overexposure and paraxial diffraction stray light effects, the paraxial portion of the image contains G and B channel color information, and also has more noise. Figure 3 The diagram shows the low threshold processing results using the traditional binarization algorithm. It can be seen that the overall brightness of the paraxial part is very high, making it impossible to perform spot edge detection and center point extraction. Figure 4 This is a schematic diagram of the high-threshold processing results under the traditional binarization algorithm. It can be observed that the loss of the far-axis spot under the high threshold is caused by the weakening of the far-axis light intensity.

[0033] (ii) Determining the zero-order center of diffraction

[0034] The transmission function of the axicon lens array of the diffraction element used in the embodiments of the present invention is as follows:

[0035] ,

[0036] where n represents the refractive index of the axicon lens array, β represents the cone base angle of the axicon lens array, and R represents the basic radius of a single axicon.

[0037] Based on the Fraunhofer diffraction intensity distribution formula described in the invention content, the light intensity distribution function on the curtain can be obtained:

[0038]

[0039] Find the coordinates of the diffraction zero-order center pixel point P0(1930, 1722), and set R1 = 600, R2 = 900, and R3 = 1700.

[0040] (III) Ring-shaped multi-channel processing and median filtering denoising

[0041] According to the pixel distance formula in the invention content, the distance d between other pixel points and the diffraction zero-order center pixel point P0 can be obtained. When d < R1, it is a pixel point in area S1; when R1 < d < R2, it is a pixel point in area S2; when R2 < d < R3, it is a pixel point in area S3; when R3 < d, it is a pixel point in area S4. Extract S1 of the B channel of the spot pattern, S2 of the G channel of the spot pattern, S3 of the R channel of the spot pattern, and S4.

[0042] (IV) Median filtering denoising and image fusion

[0043] Set a 7*7 convolution kernel and a 180-pixel threshold to perform median filtering denoising on S3, set a 7*7 convolution kernel and a 90-pixel threshold to perform median filtering denoising on S4, and perform image fusion on the obtained four ring-shaped S regions. As Figure 5 shows the schematic diagram of area division after ring-shaped multi-channel processing, Figure 6 is the schematic diagram of the complete spot gray-scale image obtained by using the ring-shaped multi-channel algorithm.

[0044] (V) Edge detection

[0045] Set edge templates for detecting the positive 45° direction and the negative 45° direction, perform two-dimensional image filtering processing, and obtain the edge contour of the spot. As Figure 7 shown is the schematic diagram of spot edge contour detection.

[0046] (VI) Construct connected regions and extract and display the spot center

[0047] After labeling the connected regions, use the existing method to measure the center of the connected regions and extract the spot center coordinates. As Figure 8The diagram shown is a schematic of the extraction of the center position of the light spot (only the coordinates of a portion of the light spot are shown).

[0048] The above description is merely a specific embodiment of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for extracting light spots from lattice structured light, comprising the following steps: Based on the intensity distribution characteristics of Fraunhofer diffraction, the zero-order center point of the collected lattice structured light was determined. Then, based on the high-brightness pixels caused by overexposure and the low-brightness pixels corresponding to the higher-order diffraction spots, three successively enlarged radius values ​​R1, R2, and R3 were determined. This divided the image into four regions, designated S1, S2, S3, and S4 from the zero-order center outwards. In region S1, the B-channel pixel values ​​were selected as the grayscale image; in region S2, the G-channel pixel values ​​were selected; and in regions S3 and S4, the R-channel pixel values ​​were selected. A high threshold is set to perform median filtering to remove noise in region S3, and a low threshold is set to perform median filtering to remove noise in region S4. The four grayscale images are superimposed to form the grayscale image of the image. An edge detection algorithm is applied to the grayscale image to determine the edge of each spot. A connected component is constructed and the center of the spot is displayed.

2. The method for extracting light spots for lattice structured light according to claim 1, characterized in that, The Fraunhofer diffraction intensity distribution described above: Where U(x,y) represents the light intensity at coordinate point (x,y), k represents the wave vector, λ represents the center wavelength of the laser, z represents the light propagation length, and t(x',y') is the transmission function of the diffraction element used, which needs to be determined according to the diffraction element used in the measurement.

3. The method for extracting light spots for lattice structured light according to claim 1, characterized in that, The formula for calculating the radius value R is as follows: Where (x0, y0) are the coordinates of the zero-order diffraction center point, and (x, y) are the coordinates of the target point.

4. The method for extracting light spots for lattice structured light according to claim 1, characterized in that, The high-threshold and low-threshold median filtering methods described above require the threshold size to be determined based on the characteristics of the image pixel size, in order to achieve better near-point and far-point denoising effects.

5. The method for extracting light spots for lattice structured light according to claim 1, characterized in that, The edge detection includes: defining two edge templates for detecting +45° and -45° directions as convolution kernels to perform convolution operations on the image to obtain the edge of the light spot.

Images