Digital three-dimensional imaging detection method and system for UV (ultraviolet) ice flower ink

Abstract

The invention discloses a digital three-dimensional imaging detection method and system for UV ice flower ink, and the method comprises the following steps: S1, irradiating the surface of to-be-detected UV ice flower ink through a multispectral illumination module, and synchronously collecting a reflected light field image set; s2, preprocessing the acquired reflected light field image set; s3, reconstructing the three-dimensional shape of the surface of the UV ice flower ink based on the preprocessed reflected light field image set; and S4, analyzing the reconstructed three-dimensional morphology through a neural network model, identifying ice flower texture defects and realizing coordinate positioning. According to the digital three-dimensional imaging detection method and system for the UV ice flower ink, the problems of sample damage, insufficient precision, low efficiency and the like in traditional detection are solved, three-dimensional reconstruction errors are reduced, time consumption of single sample detection is reduced, the defect recognition accuracy is high, and the digital three-dimensional imaging detection method and system are suitable for large-scale production quality control requirements.

Description

Technical Field

[0001] This invention relates to the field of ink quality testing technology, and in particular to a digital three-dimensional imaging testing method and system for UV ice flower ink. Background Technology

[0002] The uniformity of the micro-texture, the integrity of the crystals, and the surface defects of UV ice flower inks directly affect the appearance, texture, wear resistance, and adhesion stability of the product. They are core elements of quality control during the production process, but the current testing methods in the industry still have many technical shortcomings.

[0003] Contact roughness testers are prone to scratching the microstructure of inks during testing, and cannot penetrate into fine gaps or complex morphological areas. Furthermore, the probe diameter limits the capture of nanoscale features. Some highly elastic UV ice flower inks may also deform due to probe pressure, further leading to measurement data deviations. After the probe wears out, frequent calibration is required, which increases the testing cost and time consumption.

[0004] Conventional visual inspection can only acquire two-dimensional planar images and cannot restore three-dimensional morphological features. It is easily affected by lighting interference, resulting in grayscale distortion. When faced with the irregular crystalline texture unique to ice flower ink, it cannot present the height difference of texture undulation to judge the uniformity of crystalline layer thickness. It also lacks polarized light adaptation design, making it difficult to distinguish the texture itself from surface stains, resulting in a high defect misjudgment rate.

[0005] Manual inspection relies on the experience and judgment of operators, lacks a unified quantitative evaluation standard, and cannot accurately characterize key indicators such as crystallinity. The large differences in the threshold for judging defects among different personnel lead to poor consistency in the testing of samples in the same batch. In addition, the testing time for a single sample usually exceeds 5 minutes, and manual data recording is prone to omissions or errors, which is not conducive to quality traceability.

[0006] The general-purpose 3D measurement equipment was not optimized for the optical properties of UV ice flower ink, resulting in poor spectral matching and a 3D reconstruction error greater than 0.5. The lack of dedicated image preprocessing algorithms makes it impossible to effectively filter noise and extract texture details. Defect recognition often relies on simple threshold judgments, making it difficult to accurately classify and subdivide defect types. Furthermore, the positioning accuracy cannot meet the needs of defect repair in production. Summary of the Invention

[0007] The purpose of this invention is to provide a digital three-dimensional imaging detection method and system for UV ice flower ink. Through the collaborative design of multispectral polarization illumination, adaptive image preprocessing, high-precision three-dimensional reconstruction and intelligent defect recognition, it achieves accurate restoration of the three-dimensional morphology of the UV ice flower ink surface, quantitative evaluation of crystallinity and rapid location of defects, solving the technical pain points of existing detection methods such as sample damage, insufficient accuracy, low efficiency and strong subjectivity.

[0008] To achieve the above objectives, the present invention provides a digital three-dimensional imaging detection method for UV ice flower ink, comprising the following steps: S1. Illuminate the surface of the UV ice flower ink under test through a multispectral illumination module and simultaneously collect a set of reflected light field images; S2. Preprocess the acquired set of reflected light field images; S3. Reconstruct the three-dimensional morphology of the UV ice flower ink surface based on the preprocessed reflected light field image set; S4. Analyze the reconstructed 3D morphology using a neural network model to identify ice flower texture defects and achieve coordinate positioning.

[0009] Preferably, in S1, acquiring the reflected light field image set includes the following steps: S11. The multispectral illumination module uses a multi-angle polarization light source array, with the emitting surface of the multi-angle polarization light source array parallel to the surface of the sample to be tested. The multispectral illumination module uses a multi-angle polarization light source array, which includes 12 independently controlled LED units that are uniformly distributed in a ring. The central angle between adjacent LED units is 30°. The wavelength range of the LED units is 365nm-405nm. S12. Control the multi-angle polarization light source array to cycle through and irradiate the UV ice flower ink surface of the sample under test according to the polarization angle sequence of 0°, 45°, 90°, and 135°. Under each polarization angle illumination condition, the corresponding reflected light field image is synchronously acquired by the CCD camera through the telecentric lens; S13. Rotate the sample to be tested by an electric rotary table at a set angle, repeat S11 and S12, and obtain reflected light field images of the sample from multiple angles, which together constitute a set of reflected light field images.

[0010] Preferably, in S2, an adaptive bilateral filtering algorithm is used to perform image enhancement processing on the acquired reflected light field image set, as shown in the following formula: ; in, For the current target pixel to be processed, For target pixel neighborhood window, For target pixel neighborhood window Pixels within; For the spatial domain Gaussian kernel function, For space Gaussian kernel; For target pixel with neighboring pixels Spatial distance; Let Gaussian kernel function be the one with the range. The range (brightness) Gaussian kernel; For target pixel Pixel values ​​in the original reflected light field image For neighboring pixels Pixel values ​​in the original reflected light field image For target pixel with neighboring pixels The absolute value of the pixel value in the original reflected light field image. For enhanced target pixels The corresponding output pixel value; For the normalization coefficients to satisfy: .

[0011] Preferably, in step S2, the image enhancement processing further includes a quantitative evaluation of ice flower crystallization, with the following evaluation formula: ; in, This is an index of crystallinity, with a value ranging from 0 to 1. For the Laplace operator, The total number of pixels in the sampling area. The pixel values ​​of all pixels within the sampling area. For the sampling area The pixel value of each pixel.

[0012] Preferably, in S3, the three-dimensional morphology of the UV ice flower ink surface is reconstructed using the phase shift method, specifically as follows: Phase difference calculated using the four-step phase shift method The calculation formula is: ; in, These are the pixel coordinates of the corresponding image. , , , The grayscale value of the image after phase shift; Then the phase difference Substituting the values ​​into the height field formula to calculate the three-dimensional topography, the height field formula is: ; in, The wavelength of the light source ranges from 365nm to 405nm. The refractive index of the UV ice flower ink ranges from 1.45 to 1.55. Let be the angle of incidence of the light ray, and its value range is . .

[0013] Preferably, in S4, the neural network model uses the improved YOLOv7 architecture, adds an image preprocessing module to the input layer to normalize the image to the 0-1 range and scale it, and adds the CBAM attention mechanism to the feature extraction network. The loss function expression for the neural network model is as follows: ; in, For coordinate loss, For prospect confidence loss, For background confidence loss, For category loss, The weights for coordinate loss, The weight of the prospect confidence loss. The weights for background confidence loss. The weights for the category loss.

[0014] This invention provides a digital three-dimensional imaging inspection system for UV ice flower ink, including a multispectral illumination module, an image acquisition module, a processing and control module, and a result output module; The multispectral illumination module is used to illuminate the surface of the UV ice flower ink under test with different polarization angles; the image acquisition module and the multispectral illumination module synchronously acquire the reflected light field images of the surface of the UV ice flower ink under test under different polarization angles and different sample angles. The processing and control module is equipped with a GPU accelerator to control the collaborative work of the multispectral illumination module and the image acquisition module, and to preprocess the acquired reflected light field images; the result output module is used to generate three-dimensional topographic point cloud maps and defect analysis reports.

[0015] Preferably, the multispectral illumination module is a multi-angle polarization light source array, including 12 independently controlled LED units that are uniformly distributed in a ring, with the central angle between adjacent LED units being 30°. The wavelength range of the LED unit is 365nm-405nm, and its polarization angle is cyclically switched in four directions: 0°, 45°, 90°, and 135°.

[0016] Preferably, the image acquisition module includes a CCD camera, a telecentric lens, and an electric rotary stage for carrying and rotating the sample to be tested.

[0017] Therefore, the present invention employs the above-mentioned digital three-dimensional imaging detection method and system for UV ice flower ink, and the beneficial effects are as follows: (1) This invention realizes non-contact non-destructive testing of UV ice flower ink, avoiding the impact of contact testing on the microstructure of ink scratches or deformations. At the same time, it achieves high-precision three-dimensional morphology reconstruction through professional algorithms, clearly presenting the fine texture features. It uses a special formula to quantitatively evaluate the uniformity of crystallization, replacing the traditional qualitative judgment that relies on experience, so that the testing standards of different batches of samples are consistent, effectively solving the problems of sample damage and subjective testing.

[0018] (2) Compared with traditional manual inspection, the time consumption is greatly reduced. This invention can be smoothly adapted to the online quality control needs of large-scale production. The combination of multispectral illumination and filtering during the inspection process can effectively resist the interference of ambient light, ensure the stability and reliability of the inspection results under complex working conditions, reduce manual operation links, reduce the risk of data recording errors, and solve the problems of low efficiency and insufficient inspection stability.

[0019] (3) The present invention relies on an improved neural network model to achieve accurate identification of ice flower texture defects, which can distinguish different types of defects and clarify their location information, providing an accurate basis for subsequent defect repair.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is an overall flowchart of a digital three-dimensional imaging detection method for UV ice flower ink according to the present invention; Figure 2 This is a detailed flowchart of a digital three-dimensional imaging detection system for UV ice flower ink according to the present invention; Figure 3 This is an overall block diagram of a digital three-dimensional imaging detection system for UV ice flower ink according to the present invention; Figure 4 This is a detection result diagram of an embodiment of a digital three-dimensional imaging detection method and system for UV ice flower ink according to the present invention. Detailed Implementation

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0024] like Figure 1 As shown, a digital three-dimensional imaging detection method for UV ice flower ink includes the following steps: S1, such as Figure 2 As shown, the surface of the UV ice flower ink under test is illuminated by a multispectral illumination module, and a set of reflected light field images is acquired simultaneously. The process includes the following steps: S11. The multispectral illumination module uses a multi-angle polarized light source array, with the emitting surface of the multi-angle polarized light source array parallel to the surface of the sample to be tested. The multispectral illumination module uses a multi-angle polarized light source array, which includes 12 independently controlled LED units that are uniformly distributed in a ring. The central angle between adjacent LED units is 30°. The wavelength range of the LED units is 365nm-405nm, which is precisely matched with the spectral absorption characteristics of the UV ice flower ink.

[0025] S12. Control the multi-angle polarization light source array to cycle through and illuminate the UV ice flower ink surface of the sample under test according to the polarization angle sequence of 0°, 45°, 90°, and 135°; under the illumination conditions of each polarization angle, the CCD camera synchronously acquires the corresponding reflected light field image through the telecentric lens. The telecentric lens can eliminate perspective distortion and ensure the linear correspondence between image pixels and actual spatial positions.

[0026] S13. Rotate the sample under test within the set rotation range using an electric rotary table in 30° increments. A total of 12 sample angles were acquired. S11 and S12 were repeated to acquire reflected light field images of the sample at multiple angles, which together constituted a total of 12... A set of 48 reflected light field images ensures comprehensive coverage of sample surface information without any blind spots.

[0027] S2. An adaptive bilateral filtering algorithm is used to enhance the acquired reflected light field image set, as shown in the following formula: ; in, For the current target pixel to be processed, For 3×3 target pixels neighborhood window, For target pixel neighborhood window Pixels within; For the spatial domain Gaussian kernel function, The space Gaussian kernel has a value of 1.5. For target pixel with neighboring pixels Spatial distance.

[0028] Let Gaussian kernel function be the one with the range. The value is a Gaussian kernel for the range (brightness), with a value of 20; For target pixel Pixel values ​​in the original reflected light field image For neighboring pixels Pixel values ​​in the original reflected light field image For target pixel with neighboring pixels The absolute value of the pixel value in the original reflected light field image. For enhanced target pixels The corresponding output pixel value.

[0029] For the normalization coefficients to satisfy: .

[0030] Image enhancement processing also includes quantitative evaluation of ice flower crystallization, with the following evaluation formula: ; in, This is an index of crystallinity, with a value ranging from 0 to 1. The Laplace operator is used, with a 3×3 convolution kernel. The total number of pixels in the sampling area. No less than 10,000 pixels The pixel values ​​of all pixels within the sampling area. For the sampling area The pixel value of each pixel.

[0031] S3. Based on the preprocessed reflected light field image set, the three-dimensional morphology of the UV ice flower ink surface is reconstructed using the phase shift method, specifically: Phase difference calculated using the four-step phase shift method The calculation formula is: ; in, These are the pixel coordinates of the corresponding image. , , , The grayscale values ​​are the image values ​​after phase shifts of 0°, 45°, 90°, and 135°, respectively.

[0032] Then the phase difference Substituting the values ​​into the height field formula to calculate the three-dimensional topography, the height field formula is: ; in, The wavelength of the light source ranges from 365nm to 405nm. The refractive index of the UV ice flower ink ranges from 1.45 to 1.55. Let be the angle of incidence of the light ray, and its value range is . The error in 3D topography reconstruction is no greater than 0.1. .

[0033] S4. Analyze the reconstructed 3D morphology using a neural network model to identify ice flower texture defects and achieve coordinate positioning.

[0034] The neural network model of this invention uses an improved YOLOv7 architecture. An image preprocessing module is added to the input layer to normalize the image to the 0-1 range and scale it to 640×640 pixels. A CBAM attention mechanism is added after the C3 module in the feature extraction network.

[0035] The loss function expression for the neural network model is as follows: ; in, The CIoU loss function is used for coordinate loss. For prospect confidence loss, For background confidence loss, The cross-entropy loss function is used for class loss. =5 represents the weight of the coordinate loss. =1 represents the weight of the prospect confidence loss. =0.5 is the weight of the background confidence loss. =1 represents the weight of the category loss, and the defect localization and classification accuracy are optimized through weight allocation.

[0036] like Figure 3As shown, a digital three-dimensional imaging detection system for UV ice flower ink is used to implement the above method. It includes a multispectral illumination module, an image acquisition module, a processing and control module, and a result output module. The multispectral illumination module is used to illuminate the surface of the UV ice flower ink under test with different polarization angles. The multispectral illumination module is a multi-angle polarization light source array, including 12 independently controlled LED units uniformly distributed in a ring, with a central angle of 30° between adjacent LED units. The wavelength range of the LED units is 365nm-405nm, and their polarization angles are cyclically switched in four directions: 0°, 45°, 90°, and 135°. The switching and polarization state are synchronously controlled by the processing and control module.

[0037] The image acquisition module includes a CCD camera, a telecentric lens, and an electric rotary stage for carrying and rotating the sample under test. The CCD camera has 12 megapixels and a frame rate of [missing information]. 30fps, telecentric lens magnification of 0.5, working distance of 100mm, and rotational accuracy of the motorized rotary stage. Weigh 5kg. The image acquisition module and the multispectral illumination module simultaneously acquire images of the reflected light field on the surface of the UV ice flower ink under different polarization angles and sample angles.

[0038] The processing control module is equipped with an NVIDIA RTX 3090 GPU accelerator to control the collaborative work of the multispectral illumination module and the image acquisition module. The processing control module has a built-in collaborative control unit and an algorithm processing unit. The collaborative control unit realizes the synchronous triggering of the multispectral illumination module and the image acquisition module with a synchronization error of ≤1ms. The algorithm processing unit is used to perform preprocessing, 3D reconstruction and defect identification algorithms on the acquired reflected light field image.

[0039] The output module includes an industrial display screen and a data storage unit, which are used to generate 3D topographic point cloud maps and defect analysis reports in real time, and support export in CSV / PDF format.

[0040] Example 1 A packaging and printing company needs to conduct quality inspections on a batch of UV ice flower ink printing samples on PET substrates. The core inspection requirements are to identify three types of defects on the ink surface: pinholes, uneven texture, and incomplete crystallization, with a required inspection accuracy of 0.1. The single-sample inspection time is ≤60s, achieving non-contact, non-destructive testing to avoid sample damage. This invention's digital three-dimensional imaging inspection method and system were used to complete the inspection task.

[0041] First, select 10 UV ice flower ink printing samples with a size of 100mm×100mm. After removing surface dust, fix them in the center of the electric rotating stage and adjust the parallelism between the sample surface and the LED light-emitting surface to ≤0.5°.

[0042] After the system is started, the multispectral illumination module cycles through the polarization angle sequence. After each polarization angle illumination is completed, the CCD camera simultaneously acquires one reflected light field image. After acquiring four images with different polarization angles, the motorized rotary stage rotates 30°. The above operation is repeated until a 360° rotation is completed, and a total of 48 reflected light field images are acquired for a single sample.

[0043] The processing control module calls the adaptive bilateral filtering algorithm with a spatial Gaussian kernel of 1.5 and a range Gaussian kernel of 20 to denoise and enhance 48 images. Then, the crystallinity index of the samples is calculated using the crystallinity evaluation formula. The crystallinity of the 10 samples is between 0.82 and 0.88. The closer it is to 1, the more complete the crystallinity.

[0044] Based on the preprocessed image, a four-step phase-shifting method was used to calculate the phase difference, which was then substituted into the height field formula to complete the 3D reconstruction. The ink refractive index was set to 1.50, and the incident angle of the light rays was set to 30°. The generated 3D model point cloud density reached 1000 points / mm. 2 Reconstruction error 0.07 It clearly presents the three-dimensional undulating features of ice flower crystals on the ink surface.

[0045] An improved YOLOv7 model was used to analyze the three-dimensional morphology, successfully identifying defects in two samples, one of which contained two defects with a diameter ≤50 mm. pinhole Figure 4 As shown in (a), the coordinates are (32.5mm, 48.2mm) and (67.1mm, 29.8mm), respectively. Another piece exhibits localized areas of uneven texture, ranging from (15.3-22.6mm, 55.1-63.4mm), as shown in (a). Figure 4 As shown in (b), the defect identification accuracy is 100% and the positioning error is ≤0.1mm.

[0046] The system automatically generates a 3D topographic point cloud map and defect analysis report for each sample, clearly marking the defect type, quantity, and coordinates. The entire process of testing a single sample takes 48 seconds, meeting the production testing needs of enterprises.

[0047] Therefore, the present invention adopts the above-mentioned digital three-dimensional imaging detection method and system for UV ice flower ink, realizing non-contact non-destructive testing of UV ice flower ink samples. While ensuring detection accuracy, it greatly improves detection efficiency and effectively solves the problem of high sample damage and defect misjudgment rate in traditional testing, which is suitable for the quality control needs of large-scale production of enterprises.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for digital three-dimensional imaging inspection of UV frost ink, characterized in that, The method comprises the following steps: S1, irradiating the surface of the UV frost ink to be measured by a multi-spectral illumination module, and synchronously collecting a set of reflected light field images; S2, preprocessing the collected set of reflected light field images; S3, reconstructing the three-dimensional morphology of the surface of the UV frost ink based on the preprocessed set of reflected light field images; S4, analyzing the reconstructed three-dimensional morphology by a neural network model, identifying frost texture defects, and realizing coordinate positioning.

2. A digital 3D imaging inspection method for UV frost ink according to claim 1, characterized in that, In S1, collecting the set of reflected light field images comprises the following steps: S11, the multi-spectral illumination module selects a multi-angle polarized light source array, the light emitting surface of the multi-angle polarized light source array is parallel to the surface of the sample to be measured, the multi-angle polarized light source array selects a multi-angle polarized light source array, which comprises 12 groups of independently controlled and annularly and uniformly distributed LED units, the central angle of adjacent LED units is 30°; the wavelength range of the LED unit is 365nm-405nm; S12, controlling the multi-angle polarized light source array to cyclically switch and irradiate the surface of the sample UV frost ink at the polarization angle sequence of 0°, 45°, 90° and 135°; Under the illumination condition of each polarization angle, the corresponding reflected light field image is synchronously collected by the CCD camera through the telecentric lens; S13, rotating the sample to be measured by the electric rotating table at the set angle, repeating S11 and S12, and obtaining the reflected light field images of the sample at multiple angles to jointly constitute the set of reflected light field images.

3. A method for digital 3D imaging inspection of UV frost ink according to claim 1, characterized in that, In S2, the adaptive bilateral filtering algorithm is used for image enhancement processing of the collected set of reflected light field images, and the formula is as follows: ； wherein, is a current target pixel point to be processed, is a target pixel point a neighborhood window of the target pixel point, is a pixel point within a neighborhood window of the target pixel point a neighborhood window of the target pixel point a neighborhood window of the target pixel point; is a spatial domain Gaussian kernel function, is a spatial Gaussian kernel; is a spatial distance between the target pixel point a neighborhood pixel point a neighborhood pixel point is a value domain Gaussian kernel function, is a value domain (luminance) Gaussian kernel; is a target pixel point is a pixel value in the original reflected light field image, is a neighborhood pixel point is a pixel value in the original reflected light field image, is a target pixel point and a neighborhood pixel point is an absolute value of a pixel value in the original reflected light field image, is an enhanced target pixel point a corresponding output pixel value; for the normalization coefficients to satisfy: 。 4. A method for digital 3D imaging inspection of UV frost ink according to claim 3, characterized in that, In S2, the image enhancement processing further comprises frost crystallinity quantitative evaluation, and the evaluation formula is as follows: ； in, This is an index of crystallinity, with a value ranging from 0 to 1. For the Laplace operator, The total number of pixels in the sampling area. The pixel values ​​of all pixels within the sampling area. For the sampling area The pixel value of each pixel.

5. A method for digital 3D imaging inspection of UV frost ink according to claim 4, characterized in that, In S3, the phase shift method is used to reconstruct the three-dimensional morphology of the surface of the UV frost ink, specifically: The phase difference is calculated by a four-step phase shift method The calculation formula is: ； wherein, is a pixel coordinate of the corresponding image, , , , is the phase-shifted image gray value; The phase difference is again calculated The three-dimensional topography is calculated by substituting the height field formula, which is ； wherein, is the wavelength of the light source, having a value in the range 365nm-405nm, is the refractive index of the UV frost ink, having a value in the range 1.45-1.55, is the angle of incidence of the light, having a value in the range .

6. A method for digital 3D imaging inspection of UV frost ink according to claim 5, characterized in that, In S4, the neural network model selects an improved YOLOv7 architecture, adds an image preprocessing module to the input layer, normalizes the image to the interval of 0-1 and scales, and increases the CBAM attention mechanism in the feature extraction network; The loss function expression of the neural network model is as follows: ； wherein, is a coordinate loss, is a foreground confidence loss, is a background confidence loss, is a class loss, is a weight for the coordinate loss, is a weight for the foreground confidence loss, is a weight for the background confidence loss, is a weight for the class loss.

7. A digital three-dimensional imaging inspection system for UV frost ink for implementing the method according to any one of claims 1 to 6, characterized in that: It comprises a multi-spectral illumination module, an image acquisition module, a processing control module and a result output module; The multi-spectral illumination module is used to irradiate the surface of the UV frost ink to be measured at different polarization angles; the image acquisition module synchronously collects the corresponding reflected light field images of the surface of the UV frost ink to be measured under different polarization angles and different sample angles; The processing control module is equipped with a GPU accelerator, which is used to control the cooperative work of the multi-spectral illumination module and the image acquisition module, and to preprocess the collected reflected light field images; the result output module is used to generate a three-dimensional morphology point cloud graph and a defect analysis report.

8. A digital 3D imaging inspection system for UV frost ink according to claim 7, characterized in that: The multi-spectral illumination module is a multi-angle polarized light source array, which comprises 12 groups of independently controlled and annularly and uniformly distributed LED units, and the central angle of adjacent LED units is 30°; The wavelength range of the LED unit is 365nm-405nm, and the polarization angle is cyclically switched in four directions of 0°, 45°, 90° and 135°.

9. A digital 3D imaging inspection system for UV frost ink according to claim 7, characterized in that: The image acquisition module comprises a CCD camera, a telecentric lens and an electric rotating table for carrying and rotating the sample to be measured.

Images