Image quality enhancement method and device for image blurring caused by construction dust

Abstract

The invention discloses an image quality enhancement method and device for image blurring caused by construction dust, and belongs to the technical field of industrial image processing. The method comprises the following steps: acquiring spherical coordinates of shape point cloud data of dust particles; performing spherical harmonic transformation on the point cloud data in the spherical coordinate form based on the estimated value of the expansion coefficient, and reconstructing to obtain a dust particle model; establishing a dust particle optical scattering model; a blurred image is obtained, after color compensation is carried out on the blurred image, the blurred image is sent to a first estimation channel and a second estimation channel, the first estimation channel is used for carrying out transmission image transmission estimation on the blurred image, and the second estimation channel is used for carrying out environment optical estimation on the blurred image; and according to the transmission image transmission estimation result and the ambient light estimation result, obtaining a clear image in combination with an environment scattering model. The method has high measurement precision and efficiency, and measurement data can be utilized for multiple times, so that the cost of hardware upgrading and equipment transformation is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the field of industrial image processing, and in particular to a method and device for enhancing image quality of blurred images caused by construction dust. Background Art

[0002] Dust pollution is an increasingly prominent problem in construction, mining, road construction, and other work environments. It not only poses a serious health threat to on-site workers but also significantly negatively impacts image acquisition equipment such as video surveillance, intelligent recognition, and machine vision. Dust interference degrades the imaging quality of these devices, reducing recognition accuracy and monitoring effectiveness, posing a significant risk to project safety management.

[0003] Currently, the industry primarily employs three types of technical solutions to address dust interference. The first category involves traditional image enhancement methods, such as histogram equalization, Wiener filtering, and adaptive contrast adjustment, which primarily process the image itself. The second category involves deep learning-based approaches, including convolutional neural network dehazing algorithms, generative adversarial network image enhancement, and multi-scale feature fusion. These methods perform well in certain scenarios. The third category involves hardware improvements, such as infrared imaging, multispectral cameras, and polarization imaging systems, which can overcome dust interference to a certain extent.

[0004] However, existing technical solutions have many shortcomings. First, there is a lack of in-depth research on the geometric characteristics of dust particles and the optical scattering mechanism. Most methods ignore the shape characteristics, size distribution and scattering characteristics of dust particles in different construction environments, resulting in the established physical model being inaccurate and unable to effectively describe the mechanism of dust's impact on image quality. Secondly, the real-time performance and environmental adaptability of the algorithm are insufficient. Existing methods often have high computational complexity and slow processing speed, and are more sensitive to environmental factors such as changes in dust concentration and lighting conditions, making it difficult to meet actual engineering needs. Thirdly, the system has poor practicality. Many methods require complex parameter adjustments, have high maintenance costs, and are difficult to guarantee reliability and stability in long-term operation. Summary of the Invention

[0005] In light of the shortcomings of existing technologies, this invention provides a method and device for enhancing image quality for images blurred by construction dust. By precisely describing the geometric features and optical scattering properties of dust particles, this method addresses the technical issue of image blur caused by construction dust. This method requires no hardware modification, offers high computational efficiency, and is highly adaptable to various environments. It provides a low-cost and reliable solution for blurred image processing in engineering practice.

[0006] The technical means adopted in the present invention are as follows:

[0007] In one aspect, the present invention provides a method for enhancing image quality of images blurred by construction dust, comprising the following steps:

[0008] S1. Acquire a blurred image in a real environment, perform three-dimensional scanning on dust particles in the blurred image, obtain shape point cloud data of the dust particles, and describe the shape point cloud data of the dust particles as spherical coordinates;

[0009] S2. Under a given maximum expansion order, obtain an estimated value of the expansion coefficient by the least squares method, perform a spherical harmonic transformation on the shape point cloud data in spherical coordinate form based on the estimated value of the expansion coefficient, and reconstruct a dust particle model;

[0010] S3. Establishing a dust particle optical scattering model based on the reconstructed dust particle model, wherein the dust particle optical scattering model is used to obtain optical characteristic indicators of the dust particles according to optical parameters of the dust particles;

[0011] S4. After color compensation is performed on the blurred image, the blurred image is sent to a first estimation channel and a second estimation channel, respectively. The first estimation channel is used to estimate the medium transmittance of the blurred image, and the second estimation channel is used to estimate the ambient light intensity of the blurred image.

[0012] S5. Obtain a clear image based on the transmission estimation result of the transmission map and the ambient light estimation result in combination with an image blur degradation model. The image blur degradation model formula is:

[0013] I(x)=J(x)t(x)+A(1-t(x))

[0014] Where x represents the pixel coordinate, I(x) represents the blurred image, J(x) represents the clear image, A represents the ambient light intensity, and t(x) represents the medium transmittance map.

[0015] In one aspect, the present invention provides an image quality enhancement device for images blurred by construction dust, comprising:

[0016] a scanning unit, which is used to obtain a blurred image in a real environment, perform three-dimensional scanning on the dust particles in the blurred image, obtain shape point cloud data of the dust particles, and describe the shape point cloud data of the dust particles as spherical coordinates;

[0017] A reconstruction unit is used to obtain an estimated value of the expansion coefficient by the least square method under a given maximum expansion order, and perform a spherical harmonic transformation on the shape point cloud data in the form of spherical coordinates based on the estimated value of the expansion coefficient to reconstruct and obtain a dust particle model;

[0018] a dust particle optical scattering model construction unit, which is used to establish a dust particle optical scattering model based on the reconstructed dust particle model, wherein the dust particle optical scattering model is used to obtain optical characteristic indicators of the dust particles according to the optical parameters of the dust particles;

[0019] an estimation unit, configured to perform color compensation on the blurred image and then feed the resultant image into a first estimation channel and a second estimation channel, respectively, wherein the first estimation channel is used to estimate the medium transmittance of the blurred image and the second estimation channel is used to estimate the ambient light intensity of the blurred image;

[0020] The clarity unit is used to obtain a clear image based on the transmission estimation result of the transmission map and the ambient light estimation result in combination with the image blur degradation model. The image blur degradation model formula is:

[0021] I(x)=J(x)t(x)+A(1-t(x))

[0022] Where x represents the pixel coordinate, I(x) represents the blurred image, J(x) represents the clear image, A represents the ambient light intensity, and t(x) represents the medium transmittance map.

[0023] On the one hand, the present invention provides a storage medium, which includes a stored program, wherein when the program is run, the above-mentioned image quality enhancement method for image blurring caused by construction dust is executed.

[0024] On the other hand, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-mentioned method for enhancing image quality for images blurred by construction dust through the operation of the computer program.

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] (1) High modeling accuracy: The spherical harmonic series is used to geometrically model dust particles, breaking through the limitations of traditional spherical approximation. It can accurately describe the shape characteristics of irregular particles, making the scattering characteristic analysis more consistent with the actual situation and providing a more reliable physical basis for image quality enhancement.

[0027] (2) Strong adaptability: The method of the present invention comprehensively considers multiple influencing parameters such as the shape distribution of dust particles and environmental factors, and can adaptively adjust the treatment strategy. It is suitable for different types of construction environments and dust pollution levels and has strong generalization ability.

[0028] (3) Low cost: This method only requires upgrading the software level of the existing monitoring system, without adding or modifying hardware equipment, which greatly reduces the implementation cost. At the same time, the system is easy to maintain, operates stably and reliably, and has good engineering practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0030] Figure 1 The flowchart of an image quality enhancement method for images blurred by construction dust according to an embodiment of the present invention is shown.

[0031] Figure 2 Schematic diagram of the parameterization of the spherical surface of a dust particle with concave surface features in an embodiment of the present invention.

[0032] Figure 3 2 is a diagram of the environment scattering model architecture in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0034] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0035] Example 1

[0036] like Figure 1As shown, this embodiment provides an image quality enhancement method for images blurred by construction dust, comprising the following steps:

[0037] S1. Obtain a blurred image in a real environment, perform three-dimensional scanning on the dust particles in the blurred image, obtain shape point cloud data of the dust particles, and describe the shape point cloud data of the dust particles as spherical coordinates.

[0038] The shape point cloud data of dust particles during construction is obtained through 3D scanning technology. The point cloud data is described in the form of spherical coordinates r(φ,θ). Specifically, in the spherical coordinate system, r is the radial distance from the center point of the dust particle to the origin of the spherical coordinates, and θ is the polar angle (0≤θ≤π, the angle with the z-axis). is the azimuth ( The three together determine the spatial position of the dust particles.

[0039] Since dust particles have different shapes, simple dust particles can be directly described using spherical coordinates, but complex dust particles are difficult to describe directly. In this case, for dust particles with too complex shapes that cannot be directly described using spherical coordinates, the spherical projection method is used based on the equivalence principle of zero-genus surface and sphere in topology to project the complex shape point cloud onto the unit sphere to form a one-to-one mapping.

[0040] This embodiment preferably adopts the following steps to achieve the spherical coordinate description of any dust particle.

[0041] S101 . Classify the dust particles into first dust particles without concave surface features and second dust particles with concave surface features according to their shapes.

[0042] Optionally, S101 includes: obtaining a rotational symmetry index of the dust particle, the rotational symmetry index including the number of symmetry axes and a mirror symmetry degree index; calculating the Euler number of the particle three-dimensional grid based on the rotational symmetry index of the dust particle, thereby determining whether the dust particle has a concave surface feature.

[0043] Alternatively, S101 includes: obtaining a rotational symmetry index of the dust particle, the rotational symmetry index including the number of symmetry axes and a mirror symmetry degree index; analyzing based on the rotational symmetry index of the dust particle whether the proportion of the negative area of ​​Gaussian curvature is greater than 10%, thereby determining whether the dust particle has a concave surface feature.

[0044] S102. Describe the first dust particles directly using spherical coordinates.

[0045] S103 , based on the principle of equivalence between a zero-genus surface and a sphere in topology, a spherical projection method is used to project the point cloud of the second dust particle onto a unit sphere to form a one-to-one mapping.

[0046] Specifically, in S103, in order to solve the problem that the complex surface point cloud data of the crushed material cannot be directly described by spherical coordinates, according to the principle in topology that a closed, hole-free, zero-genus surface is topologically equivalent to a spherical surface, the existing spherical parameterization algorithm (AHSP, Advanced Hierarchical Spherical Parameterizations) is used to spherically parameterize the surface point cloud of the crushed material and map the coordinate points to the unit sphere. The AHSP spherical parameterization algorithm is used, which controls the area and length distortion, projects the closed, hole-free surface onto the unit sphere, and solves a bidirectional mapping between the complex surface points and the points on the sphere. The specific process is as follows: Figure 2 shown.

[0047] S2. Under a given maximum expansion order, an estimated value of the expansion coefficient is obtained by the least square method, and a spherical harmonic transformation is performed on the point cloud data in the form of spherical coordinates based on the estimated value of the expansion coefficient to reconstruct and obtain a dust particle model.

[0048] This step mainly uses spherical harmonic series to expand the 3D surface obtained by scanning dust particles, and calculates the estimated value of the expansion coefficient through least squares fitting, and reconstructs the 3D surface based on the estimated value of the expansion coefficient. The expansion formula is:

[0049]

[0050]

[0051] in, To reconstruct the surface, and Expansion coefficient It can be obtained by least square estimation. For example, given the maximum expansion level L max How to find the expansion coefficient Assume that the sampling points on the input surface The corresponding X coordinate is Then for the interval 1≤i≤n that satisfies According to the following formula, the following linear system can be obtained:

[0052]

[0053] in, Since n≠k in most cases, the above k )T The system can be solved by least squares fitting

[0054] because is the parameter of the original surface Estimates of (j = l 2 +l+m+1), so the original surface can be approximately reconstructed by the following formula:

[0055]

[0056] The higher the expansion level (L max The larger the value, the higher the reconstruction accuracy. and The coefficients obtained after fitting are It can be used to reconstruct the original surface with arbitrary precision.

[0057] S3. Based on the reconstructed dust particle model, a dust particle optical scattering model is established. The dust particle optical scattering model is used to obtain optical characteristic indicators of the dust particles according to the optical parameters of the dust particles.

[0058] This step measures the optical parameters of actual dust to obtain the optical properties of blasting dust particles. It then analyzes the effects of different parameter combinations and spherical harmonic frequencies in the parameterized dust particle model on the optical properties of the dust, and establishes a relationship between the spherical harmonic series parameters and the optical properties. Here, the parameter combinations in the dust particle model are different spherical harmonic series parameter combinations.

[0059] Specifically, the optical parameters of dust particles include refractive index, absorption coefficient, extinction coefficient, scattering coefficient, phase function, asymmetry factor, and particle size distribution parameters. The optical properties of dust particles are mainly characterized by indicators such as scattering angle distribution, polarization state, scattered light intensity, and absorption spectrum.

[0060] The geometric parameters of dust particles (spherical harmonic series parameters) are input into the dust particle optical scattering model to generate optical properties (such as phase function and extinction coefficient). The dust particle optical scattering model is a pre-established correspondence between spherical harmonic coefficients and optical parameters through neural network simulation. When used, the corresponding optical parameters are obtained based on the acquired spherical harmonic series parameters. Combining the optical parameters with the physical degradation model yields an image blur degradation model.

[0061] S4, after color compensation, the blurred image is sent to the first estimation channel and the second estimation channel respectively. The first estimation channel is used to estimate the medium transmission of the blurred image, and the second estimation channel is used to estimate the ambient light intensity of the blurred image. Figure 3As shown, the first estimation channel includes a minimum value filtering module, a Gaussian transformation module, and an adaptive Gaussian module; the second estimation channel includes a grayscale transformation module, a superpixel segmentation module, and an average grayscale module. Specifically:

[0062] The input data for the minimum filter module is the color-compensated blurred image. Each pixel in the blurred image contains color values ​​for three channels: red (R), green (G), and blue (B), represented as a pixel matrix. Each element in the matrix corresponds to the color value of a pixel in the image. The output data for the minimum filter module is the image after the minimum filter processing, also a pixel matrix. Each element is the processed color value of the corresponding pixel and its neighborhood.

[0063] The input data of the Gaussian transform module is the image output by the minimum filter module. The output data of the Gaussian transform module is the image after Gaussian transform, whose pixel matrix size is the same as the input image, and the color value of each pixel is Gaussian smoothed.

[0064] The input data of the adaptive Gaussian module is the image output by the Gaussian transform module. The output data of the adaptive Gaussian module is the image after the Gaussian filter is adaptively adjusted according to the local features of the image, also in the form of a pixel matrix.

[0065] The input data of the grayscale transformation module is the blurred image after color compensation. The output data of the grayscale transformation module is the grayscale image, which converts the colored blurred image into an image with only grayscale values. Each pixel is represented by a grayscale value, usually ranging from 0 to 255, and is represented as a two-dimensional matrix.

[0066] The input data of the pixel segmentation module is the grayscale image output by the grayscale transformation module. The output data of the pixel segmentation module is the result of segmenting the image into multiple superpixel regions. Each superpixel region consists of a group of adjacent pixels and can be represented as a set of multiple superpixel regions. Each region records information such as the coordinates of the pixels it contains.

[0067] The input data of the average grayscale module is the image superpixel segmentation result output by the superpixel segmentation module. The output data of the average grayscale module is the average grayscale value of each superpixel region, which is a one-dimensional array with the same number of array elements as the number of superpixel regions, and each element corresponds to the average grayscale value of a superpixel region.

[0068] First, the blurred image is color-compensated using histogram matching. Specifically, luminance channel matching is used to restore contrast. Chroma channel matching is used to suppress dust bleed. Luminance matching involves mapping the blurred image's luminance distribution (such as the V channel of HSV or grayscale values) to a reference distribution (the luminance histogram of the clear image) free of dust interference. Specifically, the cumulative distribution function (CDF) of the blurred image's luminance is calculated. The CDF of the reference image is then used to map the original luminance values ​​to a wider dynamic range, stretching pixels concentrated in the low grayscale range to a normal luminance range. This expands the pixel grayscale differences, enhancing the contrast between light and dark, thereby restoring image contrast. If dust causes an overall dark image with low contrast, matching significantly improves the distinction between dark details (such as shadows) and bright areas (such as highlights). Chroma matching involves adjusting the hue (H) and saturation (S) channels. By matching the blurred image's H / S distribution to the reference chromaticity distribution of the clear image, color cast and saturation loss caused by dust scattering are corrected. Tone correction can eliminate color casts such as yellowish and bluish that may be caused by dust, making the image color more consistent with the actual scene.

[0069] The dust particle model established based on spherical harmonic series can provide a basis for color compensation and parameter estimation. The geometric parameters of the dust particles obtained by spherical harmonic series expansion can determine optical properties such as the extinction coefficient at different wavelengths. During histogram matching, the channel weights for color compensation are adjusted based on these properties. Specifically, the extinction coefficients of the three RGB channels corresponding to the wavelength are calculated, and the inverse of the extinction coefficient is used as the initial weight. After normalization, the weights sum to 1. If the blue light extinction coefficient is the highest (the dust particles are smaller), the blue channel weight is the largest. In brightness matching, the blue light contrast is enhanced first, and in saturation adjustment, the blue saturation is restored with a higher factor.

[0070] The parameters of the image blur degradation model are then estimated using transmission map estimation and ambient light estimation. In transmission map estimation, the transmittance is calculated using information about particle shape and size as reflected by spherical harmonics combined with depth of field. In ambient light estimation, the intensity and distribution of ambient light are determined based on the particle optical scattering properties as reflected by spherical harmonics.

[0071] The calculation of transmittance t(x) is based on the dust particle shape characterized by spherical harmonic series (expansion coefficient ), particle size distribution parameters (such as equivalent diameter D) and depth of field information (such as pixel depth of field d(x)), combined with the dust scattering model, it is derived that: First, the extinction coefficient of unit volume of dust is calculated using the spherical harmonic model parameters:

[0072]

[0073] Q extis the extinction efficiency factor, which is determined by the particle shape of the spherical harmonic series expansion), and then according to the formula, the transmittance t(x)=exp(-σ ext ·d(x)), that is, the larger the depth of field, the lower the transmittance and the stronger the dust scattering attenuation.

[0074] The specific method of ambient optical estimation is as follows: the scattered light intensity distribution of dust particles is calculated using the spherical harmonic scattering model. Combined with the global statistical characteristics of the image, the average value of the top 0.1% pixels with the largest brightness value in the blurred image is taken as the ambient light intensity A. Its distribution is determined by weighted fusion of the average brightness of the region after superpixel segmentation and the scattered light intensity distribution predicted by the spherical harmonic model, that is:

[0075] A(x)=ω·mean superpixel (I(x))+(1-ω)·A·P(θ(x)),

[0076] Among them, ω is the weight coefficient, θ(x) is the scattering angle corresponding to the pixel point, and the spatial coupling estimation of the ambient light intensity and the dust scattering characteristics is realized.

[0077] S5. Obtain a clear image based on the transmission estimation result of the transmission map and the ambient light estimation result in combination with an image blur degradation model. The image blur degradation model formula is:

[0078] I(x)=J(x)t(x)+A(1-t(x))

[0079] Where x represents the pixel coordinate, I(x) represents the blurred image, J(x) represents the clear image, A represents the ambient light intensity, and t(x) represents the medium transmittance map.

[0080] This embodiment discloses a method for restoring blurred images under the influence of dust particles in a construction environment. This method is based on spherical harmonic series modeling technology and light scattering theory, and aims to study the geometric characteristics of dust particles and their interference on image quality. It specifically includes three parts: first, a spherical harmonic series model of the geometric shape of dust particles is constructed, and the spherical mapping and parameterized reconstruction of dust particles are achieved through the principle of topological homeomorphism and geometric parameterization technology; second, a scattering characteristic model is established based on the dust geometric model to analyze the scattering behavior of single particles and dust piles of different geometric shapes in a drilling and blasting construction environment; finally, in response to the degradation of imaging quality caused by dust interference, an effective image quality restoration method is proposed by integrating light scattering characteristics and restoration algorithms. This method is simple to operate, has high measurement accuracy and efficiency, and can reuse measurement data multiple times, thereby effectively reducing the cost of hardware upgrades and equipment modifications.

[0081] Example 2

[0082] This embodiment provides an image quality enhancement device for images blurred by construction dust, which is used to implement the method provided in Example 1. The device specifically includes:

[0083] a scanning unit, which is used to obtain a blurred image in a real environment, perform three-dimensional scanning on the dust particles in the blurred image, obtain shape point cloud data of the dust particles, and describe the shape point cloud data of the dust particles as spherical coordinates;

[0084] A reconstruction unit is used to obtain an estimated value of the expansion coefficient by the least square method under a given maximum expansion order, and perform a spherical harmonic transformation on the shape point cloud data in the form of spherical coordinates based on the estimated value of the expansion coefficient to reconstruct and obtain a dust particle model;

[0085] a dust particle optical scattering model construction unit, which is used to establish a dust particle optical scattering model based on the reconstructed dust particle model, wherein the dust particle optical scattering model is used to obtain optical characteristic indicators of the dust particles according to the optical parameters of the dust particles;

[0086] an estimation unit, configured to perform color compensation on the blurred image and then feed the resultant image into a first estimation channel and a second estimation channel, respectively, wherein the first estimation channel is used to estimate the medium transmittance of the blurred image and the second estimation channel is used to estimate the ambient light intensity of the blurred image;

[0087] The clarity unit is used to obtain a clear image based on the transmission estimation result of the transmission map and the ambient light estimation result in combination with the image blur degradation model. The image blur degradation model formula is:

[0088] I(x)=J(x)t(x)+A(1-t(x))

[0089] Where x represents the pixel coordinate, I(x) represents the blurred image, J(x) represents the clear image, A represents the ambient light intensity, and t(x) represents the medium transmittance map.

[0090] Example 3

[0091] This embodiment provides a storage medium, which includes a stored program. When the program is run, the image quality enhancement method for image blurring caused by construction dust given in the above embodiment 1 is executed.

[0092] Example 4

[0093] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the image quality enhancement method for image blurring caused by construction dust given in the above-mentioned embodiment 1 through the computer program.

[0094] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0095] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0097] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0098] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0099] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for enhancing image quality for images blurred by construction dust, characterized in that: The following steps are involved: S1. Acquire a blurred image in a real environment, perform three-dimensional scanning on dust particles in the blurred image, obtain shape point cloud data of the dust particles, and describe the shape point cloud data of the dust particles as spherical coordinates; S2. Under a given maximum expansion order, obtain an estimated value of the expansion coefficient by the least squares method, perform a spherical harmonic transformation on the shape point cloud data in spherical coordinate form based on the estimated value of the expansion coefficient, and reconstruct a dust particle model; S3. Establishing a dust particle optical scattering model based on the reconstructed dust particle model, wherein the dust particle optical scattering model is used to obtain optical characteristic indicators of the dust particles according to optical parameters of the dust particles; S4. After color compensation is performed on the blurred image, the blurred image is sent to a first estimation channel and a second estimation channel, respectively. The first estimation channel is used to estimate the medium transmittance of the blurred image, and the second estimation channel is used to estimate the ambient light intensity of the blurred image. S5. Obtain a clear image based on the transmission estimation result of the transmission map and the ambient light estimation result in combination with an image blur degradation model. The image blur degradation model formula is: I(x)=J(x)t(x)+A(1-t(x)) Where x represents the pixel coordinate, I(x) represents the blurred image, J(x) represents the clear image, A represents the ambient light intensity, and t(x) represents the medium transmittance map.

2. The image quality enhancement method for image blur caused by construction dust according to claim 1, characterized in that: The shape point cloud data of the dust particles are described as spherical coordinates, including: According to the shapes of the dust particles, the dust particles are divided into first dust particles without concave surface features and second dust particles with concave surface features; The first dust particles are directly described in spherical coordinate form; For the second dust particle, according to the equivalence principle of zero-genus surface and sphere in topology, a spherical projection method is used to project the second dust particle point cloud onto the unit sphere to form a one-to-one mapping.

3. The image quality enhancement method for image blur caused by construction dust according to claim 2, characterized in that: According to the shape of the dust particles, the dust particles are divided into first dust particles without concave surface features and second dust particles with concave surface features, including: Obtaining a rotational symmetry index of the dust particles, wherein the rotational symmetry index includes an index of the number of symmetry axes and a degree of mirror symmetry; The Euler number of the three-dimensional mesh of the dust particles is calculated based on the rotational symmetry index of the dust particles to determine whether the dust particles have concave surface features.

4. The image quality enhancement method for image blur caused by construction dust according to claim 2, characterized in that: According to the shape of the dust particles, the dust particles are divided into first dust particles without concave surface features and second dust particles with concave surface features, including: Obtaining a rotational symmetry index of the dust particles, wherein the rotational symmetry index includes an index of the number of symmetry axes and a degree of mirror symmetry; Based on the rotational symmetry index of the dust particles, the proportion of the negative Gaussian curvature area is analyzed to see whether it is greater than 10%, thereby determining whether the dust particles have concave surface characteristics.

5. The image quality enhancement method for image blur caused by construction dust according to claim 1, characterized in that: The optical parameters of the dust particles include refractive index, absorption coefficient, extinction coefficient, scattering coefficient, phase function, asymmetry factor and particle size distribution parameters; the optical characteristic indicators of the dust particles include scattering angle distribution, polarization state, scattered light intensity and absorption spectrum.

6. The image quality enhancement method for image blur caused by construction dust according to claim 1, characterized in that: The first estimation channel includes a minimum filter module, a Gaussian transform module and an adaptive Gaussian module; The second estimation channel includes a grayscale transformation module, a superpixel segmentation module and an average grayscale module.

7. An image quality enhancement device for images blurred by construction dust, characterized in that: include: a scanning unit, which is used to obtain a blurred image in a real environment, perform three-dimensional scanning on the dust particles in the blurred image, obtain shape point cloud data of the dust particles, and describe the shape point cloud data of the dust particles as spherical coordinates; A reconstruction unit is used to obtain an estimated value of the expansion coefficient by the least square method under a given maximum expansion order, and perform a spherical harmonic transformation on the shape point cloud data in the form of spherical coordinates based on the estimated value of the expansion coefficient to reconstruct and obtain a dust particle model; a dust particle optical scattering model construction unit, which is used to establish a dust particle optical scattering model based on the reconstructed dust particle model, wherein the dust particle optical scattering model is used to obtain optical characteristic indicators of the dust particles according to the optical parameters of the dust particles; an estimation unit, configured to perform color compensation on the blurred image and then feed the resultant image into a first estimation channel and a second estimation channel, respectively, wherein the first estimation channel is used to estimate the medium transmittance of the blurred image and the second estimation channel is used to estimate the ambient light intensity of the blurred image; The clarity unit is used to obtain a clear image based on the transmission estimation result of the transmission map and the ambient light estimation result in combination with the image blur degradation model. The image blur degradation model formula is: I(x)=J(x)t(x)+A(1-t(x)) Where x represents the pixel coordinate, I(x) represents the blurred image, J(x) represents the clear image, A represents the ambient light intensity, and t(x) represents the medium transmittance map.

8. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is run, the image quality enhancement method for image blurring caused by construction dust as described in any one of claims 1 to 6 is executed.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the image quality enhancement method for image blurring caused by construction dust according to any one of claims 1 to 6 by running the computer program.

Images