A method, device, equipment and medium for suppressing hologram speckle noise

Abstract

The present invention discloses a method, device, equipment and medium for suppressing speckle noise of a hologram, wherein the method comprises the following steps: performing frequency screening and edge enhancement on an image to be processed; performing edge detection on an edge-enhanced image based on an edge detection operator and calculating an operator value; calculating multiple gradients of the image to be processed in different directions, calculating multiple diffusion coefficients according to an upper threshold of edge detection, and obtaining a divergence operator according to the gradient and the diffusion coefficient; performing noise suppression on the image to be processed according to the operator value and the divergence operator to obtain a target image; shortening the edge detection time by highlighting the edge detail information of the image; reducing the number of cyclic calculations in the speckle noise suppression process of the hologram, and making the denoised image clearer; and improving the denoising efficiency and denoising accuracy.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and in particular to a method, device, equipment and medium for suppressing hologram speckle noise. Background Art

[0002] At present, the methods for suppressing hologram speckle noise can be divided into two categories. One is to do it from the hardware perspective, replacing the high-coherence laser light source with a low-coherence light source, such as an LED light source, or physically removing the high coherence of the laser, such as using a double rotating plate to decoherence. The other is based on image processing algorithms, such as suppressing hologram speckle noise based on sine-cosine decomposition combined with filtering algorithms. However, the current methods for suppressing hologram speckle noise still need to be improved in terms of denoising efficiency and denoising accuracy. Summary of the invention

[0003] The purpose of the present invention is to solve at least one of the technical problems existing in the prior art and to provide a method, device, equipment and medium for suppressing hologram speckle noise.

[0004] The technical solution adopted by the present invention to solve the problem is:

[0005] A first aspect of the present invention is a method for suppressing speckle noise of a hologram, comprising:

[0006] Perform frequency screening on the image to be processed, and use the screened high-frequency components to reconstruct the image to obtain an edge-enhanced image;

[0007] Performing edge detection on the edge-enhanced image based on an edge detection operator to obtain an edge detection result, and obtaining an operator value according to the value of the edge detection result, wherein both the edge detection result and the operator value correspond one-to-one to element points of the edge-enhanced image;

[0008] For each pixel point of the image to be processed, multiple gradients of the image to be processed in different directions are calculated, multiple diffusion coefficients are calculated according to an upper threshold of edge detection, the multiple diffusion coefficients correspond to the multiple gradients in one-to-one direction, and a divergence operator is obtained according to the gradients and the diffusion coefficients;

[0009] According to the operator value and the divergence operator, noise suppression is performed on the image to be processed to obtain a target image.

[0010] According to the first aspect of the present invention, the method of performing frequency screening on the image to be processed and reconstructing the image using the screened high-frequency components to obtain an edge-enhanced image includes:

[0011] Performing modal decomposition on the image to be processed until the intrinsic modal function can no longer be extracted, obtaining a plurality of intrinsic modal functions from high frequency to low frequency, wherein the intrinsic modal function contains the high frequency component of the image to be processed, and the next intrinsic modal function is obtained based on the previous intrinsic modal function;

[0012] Image reconstruction is performed according to the intrinsic mode function to obtain the edge enhanced image.

[0013] According to the first aspect of the present invention, the modal decomposition of the image to be processed is performed until the intrinsic modal function can no longer be extracted to obtain a plurality of intrinsic modal functions from high frequency to low frequency, including:

[0014] First decomposition sub-step: determining the maximum and minimum of the image to be processed, and constructing a maximum surface and a minimum surface according to the maximum and the minimum, and taking the difference between the mean of the maximum surface and the minimum surface and the image to be processed as the first difference;

[0015] Second decomposition sub-step: taking the first difference as the input of the first decomposition sub-step to iterate until an iteration stop condition is met, and outputting an iterative result of the intrinsic mode function;

[0016] The difference between the input of the first decomposition sub-step and the iterative result is used as the input of the first decomposition sub-step, and the first decomposition sub-step and the second decomposition sub-step are repeatedly executed until the intrinsic mode function can no longer be extracted, thereby obtaining multiple intrinsic mode functions from high frequency to low frequency.

[0017] According to the first aspect of the present invention, the step of performing edge detection on the edge enhanced image based on an edge detection operator to obtain an edge detection result includes:

[0018] Performing image smoothing processing on the edge enhanced image by using a filter to obtain a smoothed image;

[0019] Calculating the magnitude and direction of the gradient of the smoothed image;

[0020] Perform non-maximum suppression on the gradient amplitude according to the direction of the gradient;

[0021] Set the upper and lower thresholds and use the double threshold method to determine the potential boundary;

[0022] Based on hysteresis edge tracking, edge detection results are determined from potential boundaries.

[0023] According to a first aspect of the present invention, the operator value comprises a first operator value and a second operator value different from the first operator value;

[0024] The obtaining an operator value according to the value of the edge detection result includes:

[0025] When the edge detection result satisfies a first value condition, outputting the first operator value;

[0026] When the edge detection result satisfies a second value condition, the second operator value is output.

[0027] According to the first aspect of the present invention, obtaining a divergence operator according to the gradient and the diffusion coefficient comprises:

[0028] The gradients and diffusion coefficients in the same direction are multiplied one by one, and then the sum of the products of multiple gradients and diffusion coefficients is multiplied by a weight coefficient to obtain the divergence operator.

[0029] According to the first aspect of the present invention, the step of performing noise suppression on the image to be processed according to the operator value and the divergence operator to obtain a target image comprises:

[0030] Obtaining a correction value according to the operator value and the divergence operator;

[0031] The image to be processed is added to the correction value to obtain the target image.

[0032] A second aspect of the present invention is a device for suppressing hologram speckle noise, comprising:

[0033] The edge enhancement unit is used to perform frequency screening on the image to be processed, and reconstruct the image using the screened high-frequency components to obtain an edge-enhanced image;

[0034] An edge detection unit, configured to perform edge detection on the edge enhanced image based on an edge detection operator to obtain an edge detection result, and obtain an operator value according to the value of the edge detection result, wherein both the edge detection result and the operator value correspond one-to-one to element points of the edge enhanced image;

[0035] A divergence operator calculation unit, used to calculate multiple gradients of the image to be processed in different directions for each pixel point of the image to be processed, calculate multiple diffusion coefficients according to an upper threshold of edge detection, the multiple diffusion coefficients correspond to the multiple gradients in one-to-one direction, and obtain a divergence operator according to the gradients and the diffusion coefficients;

[0036] The noise suppression unit is used to suppress the noise of the image to be processed according to the operator value and the divergence operator to obtain a target image.

[0037] According to a third aspect of the present invention, an electronic device comprises:

[0038] at least one memory;

[0039] at least one processor;

[0040] at least one computer program;

[0041] The computer program is stored in the memory, and the processor executes the at least one computer program to implement the method for suppressing hologram speckle noise as described in the first aspect of the present invention.

[0042] A fourth aspect of the present invention is a computer-readable storage medium storing computer-executable instructions for causing a computer to execute the method for suppressing hologram speckle noise as described in the first aspect of the present invention.

[0043] The above scheme has at least the following beneficial effects: by highlighting the edge detail information of the image, the edge detection time can be shortened; then, according to the operator value and the divergence operator of the edge detection, the image to be processed is filtered and denoised, so that the number of loop calculations in the speckle noise suppression process of the hologram is reduced, and the output denoised image is clearer; the denoising efficiency and denoising accuracy are improved.

[0044] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The present invention will be further described below in conjunction with the accompanying drawings and examples.

[0046] Figure 1 is a flow chart of a method for suppressing hologram speckle noise according to an embodiment of the present invention;

[0047] Figure 2 is a flowchart of step S100;

[0048] Figure 3 is a flowchart of step S200;

[0049] Figure 4 is a structural diagram of a device for suppressing hologram speckle noise according to an embodiment of the present invention;

[0050] Figure 5 It is a schematic diagram of the directions corresponding to the four gradients of the pixel points of the image to be processed;

[0051] Figure 6 is a schematic diagram of the image to be processed;

[0052] Figure 7 Schematic diagram of a target image processed by the hologram speckle noise suppression method. DETAILED DESCRIPTION

[0053] This section will describe in detail the specific embodiments of the present invention. The preferred embodiments of the present invention are shown in the accompanying drawings. The purpose of the accompanying drawings is to supplement the description of the text part of the specification with graphics, so that people can intuitively and vividly understand each technical feature and the overall technical solution of the present invention, but it cannot be understood as a limitation on the scope of protection of the present invention.

[0054] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0055] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0056] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0057] Holograms are based on the wave principle of light, and are formed by the interference of object light (light carrying object information) and reference light (light without object information) and recorded by CCD / CMOS. Due to the high coherence of lasers, when lasers are irradiated on rough surfaces, they scatter, forming speckle noise in the hologram.

[0058] An embodiment of the first aspect of the present invention provides a method for suppressing speckle noise of a hologram.

[0059] Reference Figure 1 , the methods for suppressing hologram speckle noise include:

[0060] Step S100, performing frequency screening on the image to be processed, and reconstructing the image using the screened high-frequency components to obtain an edge-enhanced image;

[0061] Step S200: performing edge detection on the edge enhanced image based on the edge detection operator to obtain an edge detection result, and obtaining an operator value according to the value of the edge detection result, wherein both the edge detection result and the operator value correspond one-to-one to the element points of the edge enhanced image;

[0062] Step S300, for each pixel point of the image to be processed, multiple gradients of the image to be processed in different directions are calculated, multiple diffusion coefficients are calculated according to the upper threshold of edge detection, the multiple diffusion coefficients correspond to the multiple gradients in one-to-one direction, and a divergence operator is obtained according to the gradient and the diffusion coefficient;

[0063] Step S400: noise suppression is performed on the image to be processed according to the operator value and the divergence operator to obtain a target image.

[0064] Reference Figure 2 , corresponding to step S100, the image to be processed is subjected to frequency screening, and the screened high-frequency components are used to reconstruct the image to obtain an edge-enhanced image, including but not limited to the following steps:

[0065] Step S110, performing modal decomposition on the image to be processed until no intrinsic modal function can be extracted, and obtaining a plurality of intrinsic modal functions from high frequency to low frequency, wherein the intrinsic modal function contains the high frequency component of the image to be processed, and the next intrinsic modal function is obtained based on the previous intrinsic modal function;

[0066] Step S120: reconstruct the image according to the intrinsic mode function to obtain an edge enhanced image.

[0067] Among them, for step S110, modal decomposition is performed on the image to be processed until the intrinsic modal function can no longer be extracted, and multiple intrinsic modal functions from high frequency to low frequency are obtained, including but not limited to the following steps:

[0068] The first decomposition sub-step: determining the maximum and minimum values ​​of the image to be processed, and constructing a maximum surface and a minimum surface according to the maximum and minimum values, and taking the difference between the mean of the maximum surface and the minimum surface and the image to be processed as the first difference;

[0069] Second decomposition sub-step: taking the first difference as the input of the first decomposition sub-step to iterate until the iteration stop condition is met, and outputting the iterative result of the intrinsic mode function;

[0070] The difference between the input of the first decomposition sub-step and the iterative result is used as the input of the first decomposition sub-step, and the first decomposition sub-step and the second decomposition sub-step are repeatedly executed until the intrinsic mode function can no longer be extracted, and multiple intrinsic mode functions from high frequency to low frequency are obtained.

[0071] Specifically, for step S110, for the image to be processed, the corresponding function is f(x, y), then the maximum and minimum values ​​of the image to be processed are the maximum and minimum values ​​of f(x, y). The maximum surface u(x, y) is constructed according to the maximum value by interpolation, and the minimum surface v(x, y) is constructed according to the minimum value by interpolation. Then the maximum surface and the minimum surface are the upper and lower envelope surfaces. The mean of the upper and lower envelope surfaces is: The difference between the mean and the image to be processed is taken as the first difference h1(x,y), then h1(x,y)=f(x,y)-e1(x,y).

[0072] Replace f(x, y) with h1(x, y) as the new input of the first decomposition sub-step, and repeat the first decomposition sub-step k times until the iteration stop condition is met. Get the k-times iteration result h 1k (x, y); the value range of SD is generally between 0.1-0.5. 1k (x, y) is an intrinsic mode function, which is the image function component with the highest frequency containing the image f(x, y), denoted as C1(x, y); C1(x, y) is the first intrinsic mode function.

[0073] The difference between the image to be processed and the iterative result is recorded as r1(x,y), that is, r1(x,y)=f(x,y)-C1(x,y); r1(x,y) is used as the input of the first decomposition sub-step, and the first decomposition sub-step and the second decomposition sub-step are repeated to obtain the second intrinsic mode function, recorded as C2(x,y).

[0074] Then r2(x,y)=f(x,y)-C1(x,y)-C2(x,y); take r2(x,y) as the input of the first decomposition sub-step, repeat the first decomposition sub-step and the second decomposition sub-step to obtain the third intrinsic mode function, denoted as C3(x,y).

[0075] Repeat the above process, when r n (x,y) and C n (x,y) is less than the preset error or r n When (x, y) is a monotonic function, it is no longer possible to extract the intrinsic mode function from the image to be processed. n (x,y) represents the average trend of the slowly changing image.

[0076] Then the image to be processed can be represented by n intrinsic mode functions;

[0077] For step S120 , the detail information of the image to be processed is mainly concentrated on the high-frequency component C1 , and the image is reconstructed through C1 to obtain an edge-enhanced image f′(x, y), then f′(x, y)=C1 .

[0078] Reference Figure 3 In some embodiments of the present invention, in step S200, edge detection is performed on the edge enhanced image based on an edge detection operator to obtain an edge detection result, including but not limited to the following steps:

[0079] Step S210, performing image smoothing processing on the edge enhanced image through a filter to obtain a smoothed image;

[0080] Step S220, calculating the magnitude and direction of the gradient of the smoothed image;

[0081] Step S230, performing non-maximum suppression on the gradient amplitude according to the direction of the gradient;

[0082] Step S240, setting an upper threshold and a lower threshold, and using a double threshold method to determine a potential boundary;

[0083] Step S250: Determine edge detection results from potential boundaries based on hysteresis edge tracking.

[0084] In step S210, the edge enhanced image is smoothed by using a Gaussian filter to remove noise with large grayscale changes in the image, thereby obtaining a smoothed image.

[0085] For step S220, the image difference Sobel operator, Prewitt operator or Roberts operator is used to calculate the difference G between the pixels in the X direction and the Y direction in the image. x and G y , then the gradient amplitude G can be expressed as: or G=|G x |+|G y |; The direction of the gradient θ can be expressed as

[0086] As for step S230, the purpose of step S230 is to make the blurred boundary clear. In layman's terms, it is to retain the maximum value of the gradient amplitude at each pixel point and delete other values.

[0087] For each pixel, perform the following operations: where the range of θ is (-π, π), which can be calculated as and Approximate to the horizontal (0°), vertical (90°) and two diagonal (45° and 135°) directions; compare the gradient amplitude of the pixel point with the pixels in the positive and negative directions of its gradient direction. If the pixel point has the largest gradient amplitude, it is retained, otherwise it is deleted.

[0088] For step S240, there are still many noise points in the image after non-maximum suppression. Set an upper threshold and a lower threshold (usually specified manually). If the pixel in the image is greater than the upper threshold, it is considered to be a boundary (called a strong boundary) and is retained; if it is less than the lower threshold, it is considered not to be a boundary and is deleted; if it is between the upper threshold and the lower threshold, it is considered to be a candidate (called a weak boundary) and needs to be further processed.

[0089] For step S250, the strong boundary can be considered as a real boundary, while the weak boundary may be a real boundary or a false boundary caused by noise or color change. It is generally believed that the weak boundary caused by the real boundary is connected to the strong boundary, while the weak boundary points caused by noise are not, so it is necessary to perform hysteresis edge tracking. The hysteresis edge tracking algorithm detects the 8-connected area pixels of a weak boundary. The weak boundary connected to the strong boundary is considered to be the boundary, and other weak boundaries are suppressed.

[0090] In step S200, the operator value includes a first operator value and a second operator value different from the first operator value; specifically, the first operator value is 0.99, and the second operator value is 0.01. The edge detection result C(x, y) actually has only two cases: 0 and 1.

[0091] Obtaining the operator value according to the value of the edge detection result includes but is not limited to the following steps:

[0092] When the edge detection result satisfies the first value condition C(x, y)=1, the first operator value K=0.99 is output;

[0093] When the edge detection result satisfies the second value condition C(x, y)=0, the second operator value K=0.01 is output.

[0094] In some embodiments of the present invention, for step S300, multiple gradients of the image to be processed in different directions are calculated for each pixel point of the image to be processed.

[0095] Reference Figure 5 , Figure 5 It is a schematic diagram of the directions corresponding to the four gradients of the pixel points of the image to be processed. It should be noted that the four directions of up, down, left, and right are usually taken, that is, for each pixel point, four gradients are calculated, which are and

[0096] Among them, multiple diffusion coefficients are calculated according to the upper threshold of edge detection. Calculate the four diffusion coefficients c of the pixel N 、c S 、c W and c EThe four diffusion coefficients correspond to the four gradients in one-to-one direction. c is the diffusion coefficient, k is the upper threshold of edge detection, and x is the absolute value of the difference between the pixel and the adjacent pixel.

[0097] The divergence operator is obtained according to the gradient and the diffusion coefficient. Specifically, the gradient and the diffusion coefficient in the same direction are multiplied one by one, and then the sum of the products of multiple gradients and diffusion coefficients is multiplied by the weight coefficient to obtain the divergence operator. The weight coefficient can be That is, the divergence operator can be expressed as:

[0098] In some embodiments of the present invention, for step S400, noise suppression is performed on the image to be processed according to the operator value and the divergence operator to obtain a target image, including but not limited to the following steps:

[0099] According to the operator value and the divergence operator, the correction value is obtained, and the correction value can be expressed as: (1-K)div;

[0100] The target image is obtained by adding the image to be processed H'(x, y) and the correction value. In fact, the hologram H'(x, y) obtained by the optical recording system contains the information of the hologram itself H(x, y) and the speckle noise N(x, y); that is, the target image can be expressed as: H"(x, y) = H'(x, y) + (1-K)div.

[0101] In this embodiment, by highlighting the edge detail information of the image, the edge detection time can be shortened; then, according to the operator value of the edge detection and the divergence operator, the image to be processed is filtered and denoised, so that the number of loop calculations is reduced during the speckle noise suppression process of the hologram, and the output denoised image is clearer, thereby improving the denoising efficiency and denoising accuracy.

[0102] PSNR, SSIM, EPI and SSI were used as indicators for quantitative analysis.

[0103] Among them, the calculation formula of PSNR is: n is 8.

[0104] MSE is the mean square error between the original and processed images; the calculation formula for MSE is: K(x,y) is the original image to be processed; I(x,y) is the target image after noise suppression processing.

[0105] The calculation formula of SSIM is: SSIM = l(x,y)·c(x,y)·s(x,y). Among them, l(x,y) is the brightness of the image; c(x,y) is the contrast of the image; s(x,y) is the structure of the image.

[0106] have

[0107] When the values ​​of PSNR and SSIM are larger, the image distortion is smaller and the effect is better.

[0108] The calculation formula of EPI is: in, and Indicates that the image suppresses the first two adjacent pixels. and Represents two adjacent pixels after image suppression.

[0109] The calculation formula for SSI is: Among them, I after represents the image after speckle suppression, I before represents the image before speckle suppression, i.e., the noisy image, Represents the standard deviation of image I(x,y), μ I Represents the mean of the image I(x,y).

[0110] The larger the EPI value, the better the details of the original image are preserved after processing. The smaller the SSI value, the better the filtering effect.

[0111] Reference Figure 6 and Figure 7 , Figure 6 is the image to be processed; Figure 7 Is Figure 6 The target image is obtained by processing the image to be processed in the hologram by using the hologram speckle noise suppression method.

[0112] An embodiment of the second aspect of the present invention provides a device for suppressing speckle noise of a hologram.

[0113] Reference Figure 4 The device for suppressing hologram speckle noise includes an edge enhancement unit 10, an edge detection unit 20, a divergence operator calculation unit 30 and a noise suppression unit 40.

[0114] Among them, the edge enhancement unit 10 is used to perform frequency screening on the image to be processed, and reconstruct the image using the screened high-frequency components to obtain an edge enhanced image; the edge detection unit 20 is used to perform edge detection on the edge enhanced image based on the edge detection operator to obtain an edge detection result, and obtain the operator value according to the value of the edge detection result, and the edge detection result and the operator value both correspond one-to-one to the element points of the edge enhanced image; the divergence operator calculation unit 30 is used to calculate multiple gradients of the image to be processed in different directions for each pixel point of the image to be processed, and calculate multiple diffusion coefficients according to the upper threshold of the edge detection, the multiple diffusion coefficients correspond one-to-one to the multiple gradients in the direction, and the divergence operator is obtained according to the gradient and the diffusion coefficient; the noise suppression unit 40 is used to suppress the noise of the image to be processed according to the operator value and the divergence operator to obtain the target image.

[0115] It should be noted that the various units of the noise suppression device adopted in the embodiment of the second aspect of the present invention correspond one-to-one to the various steps of the noise suppression method adopted in the embodiment of the first aspect of the present invention, and both have the same technical solution, solve the same technical problems, and bring the same technical effects. Therefore, the passenger flow prediction device will not be described in detail one by one.

[0116] An embodiment of the third aspect of the present invention provides an electronic device. The electronic device includes at least one memory; at least one processor; and at least one computer program. The computer program is stored in the memory, and the processor executes the at least one computer program to implement the method for suppressing hologram speckle noise as described in the embodiment of the first aspect of the present invention.

[0117] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0118] An embodiment of the fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method for suppressing hologram speckle noise as described in the embodiment of the first aspect of the present invention.

[0119] It will be appreciated by those skilled in the art that all or some of the steps and systems in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium and a communication medium. As known to those skilled in the art, the term computer storage medium includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, or other memory technologies, CD-ROM, digital versatile disks, or other optical disk storage, magnetic tape, disk storage, or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. In addition, it is known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium.

[0120] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation mode. As long as the technical effect of the present invention is achieved by the same means, it should belong to the protection scope of the present invention.

Claims

1. A method for suppressing speckle noise of a hologram, characterized in that: include: Perform frequency screening on the image to be processed, and use the screened high-frequency components to reconstruct the image to obtain an edge-enhanced image; Performing edge detection on the edge-enhanced image based on an edge detection operator to obtain an edge detection result, and obtaining an operator value according to the value of the edge detection result, wherein both the edge detection result and the operator value correspond one-to-one to element points of the edge-enhanced image; For each pixel point of the image to be processed, multiple gradients of the image to be processed in different directions are calculated, multiple diffusion coefficients are calculated according to an upper threshold of edge detection, the multiple diffusion coefficients correspond to the multiple gradients in one-to-one direction, and a divergence operator is obtained according to the gradients and the diffusion coefficients; According to the operator value and the divergence operator, noise suppression is performed on the image to be processed to obtain a target image; The operator value includes a first operator value and a second operator value different from the first operator value; and obtaining the operator value according to the value of the edge detection result includes: When the edge detection result satisfies the first value condition C(x, y)=1, the first operator value is output as 0.99; when the edge detection result satisfies the second value condition C(x, y)=0, the second operator value is output as 0.01; C(x, y) is the edge detection result; The diffusion coefficient is based on The calculated value is: c is the diffusion coefficient, k is the upper threshold of edge detection, and x is the absolute value of the difference between the pixel and its adjacent pixels.

2. The method for suppressing hologram speckle noise according to claim 1, characterized in that: The method of performing frequency screening on the image to be processed and reconstructing the image using the screened high-frequency components to obtain an edge-enhanced image includes: Performing modal decomposition on the image to be processed until no intrinsic modal function can be extracted, and obtaining a plurality of intrinsic modal functions from high frequency to low frequency, wherein the intrinsic modal function includes the high frequency component of the image to be processed, and the next intrinsic modal function is obtained based on the previous intrinsic modal function; Image reconstruction is performed according to the intrinsic mode function to obtain the edge enhanced image.

3. The method for suppressing hologram speckle noise according to claim 2, characterized in that: The modal decomposition of the image to be processed is performed until the intrinsic modal function can no longer be extracted, and a plurality of intrinsic modal functions from high frequency to low frequency are obtained, including: First decomposition sub-step: determining the maximum and minimum of the image to be processed, and constructing a maximum surface and a minimum surface according to the maximum and the minimum, and taking the difference between the mean of the maximum surface and the minimum surface and the image to be processed as the first difference; Second decomposition sub-step: taking the first difference as the input of the first decomposition sub-step to iterate until an iteration stop condition is met, and outputting an iterative result of the intrinsic mode function; The difference between the input of the first decomposition sub-step and the iterative result is used as the input of the first decomposition sub-step, and the first decomposition sub-step and the second decomposition sub-step are repeatedly executed until the intrinsic mode function can no longer be extracted, thereby obtaining multiple intrinsic mode functions from high frequency to low frequency.

4. The method for suppressing hologram speckle noise according to claim 1, characterized in that: The performing edge detection on the edge enhanced image based on the edge detection operator to obtain the edge detection result includes: Performing image smoothing processing on the edge enhanced image by using a filter to obtain a smoothed image; Calculating the magnitude and direction of the gradient of the smoothed image; Perform non-maximum suppression on the gradient amplitude according to the direction of the gradient; Set the upper and lower thresholds and use the double threshold method to determine the potential boundary; Based on hysteresis edge tracking, edge detection results are determined from potential boundaries.

5. The method for suppressing hologram speckle noise according to claim 1, characterized in that: The obtaining a divergence operator according to the gradient and the diffusion coefficient comprises: The gradients and diffusion coefficients in the same direction are multiplied one by one, and then the sum of the products of multiple gradients and diffusion coefficients is multiplied by a weight coefficient to obtain the divergence operator.

6. The method for suppressing hologram speckle noise according to claim 1, characterized in that: The step of performing noise suppression on the image to be processed according to the operator value and the divergence operator to obtain a target image includes: Obtaining a correction value according to the operator value and the divergence operator; The image to be processed is added to the correction value to obtain the target image.

7. A device for suppressing speckle noise of a hologram, characterized in that: include: The edge enhancement unit is used to perform frequency screening on the image to be processed, and reconstruct the image using the screened high-frequency components to obtain an edge-enhanced image; An edge detection unit, configured to perform edge detection on the edge enhanced image based on an edge detection operator to obtain an edge detection result, and obtain an operator value according to the value of the edge detection result, wherein both the edge detection result and the operator value correspond one-to-one to element points of the edge enhanced image; A divergence operator calculation unit, used to calculate multiple gradients of the image to be processed in different directions for each pixel point of the image to be processed, calculate multiple diffusion coefficients according to an upper threshold of edge detection, the multiple diffusion coefficients correspond to the multiple gradients in one-to-one direction, and obtain a divergence operator according to the gradients and the diffusion coefficients; A noise suppression unit, configured to perform noise suppression on the image to be processed according to the operator value and the divergence operator to obtain a target image; The operator value includes a first operator value and a second operator value different from the first operator value; and obtaining the operator value according to the value of the edge detection result includes: When the edge detection result satisfies the first value condition C(x, y)=1, the first operator value is output as 0.99; when the edge detection result satisfies the second value condition C(x, y)=0, the second operator value is output as 0.01; C(x, y) is the edge detection result; The diffusion coefficient is based on The calculated value is: c is the diffusion coefficient, k is the upper threshold of edge detection, and x is the absolute value of the difference between the pixel and its adjacent pixels.

8. An electronic device, characterized in that: include: at least one memory; at least one processor; at least one computer program; The computer program is stored in the memory, and the processor executes the at least one computer program to implement the method for suppressing hologram speckle noise according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method for suppressing hologram speckle noise according to any one of claims 1 to 6.

Images