An image denoising method and system

Abstract

Embodiments of the present specification provide an image denoising method and system, the method comprising decomposing a first image and a second image respectively to obtain a first image output result and a second image output result; dividing the first image output result into a plurality of first image blocks according to a preset block mode, wherein the preset block mode is determined according to the first image output result; determining, based on the plurality of first image blocks, a plurality of second image blocks matched with the plurality of first image blocks from the second image output result, and determining, based on the plurality of first image blocks and the plurality of second image blocks, a displacement amount of at least one second image block in the plurality of second image blocks relative to a matched first image block; shifting the plurality of second image blocks based on the displacement amount, and superimposing and denoising the shifted second image and the first image.

Description

TECHNICAL FIELD

[0001] The present specification relates to the technical field of medical technology, and in particular to an image denoising method and system. BACKGROUND

[0002] In medical image examination, X-ray technology is one of the most widely used technologies. For real-time fluoroscopy images (i.e., X-ray images), a recursive or multi-frame image superposition method is usually used to reduce image noise, but when there are moving objects in the image (e.g., various autonomous and non-autonomous movements of the examination object), multi-frame superposition will bring obvious motion artifacts, and less or no superposition will cause large noise in the image.

[0003] Therefore, it is desirable to provide an image denoising method and system. SUMMARY

[0004] One of the embodiments of the present specification provides an image denoising method. The method comprises: decomposing a first image and a second image respectively to obtain a first image output result and a second image output result; dividing the first image output result into a plurality of first image blocks according to a preset block division manner, the preset block division manner being determined according to the first image output result; determining a plurality of second image blocks matched with the plurality of first image blocks from the second image output result based on the plurality of first image blocks, and determining a displacement amount of at least one of the second image blocks relative to the matched first image block in the plurality of second image blocks based on the plurality of first image blocks and the plurality of second image blocks; shifting the plurality of second image blocks based on the displacement amount, and superimposing and denoising the shifted second image and the first image.

[0005] One of the embodiments of the present specification provides an image denoising system, comprising an image decomposition module, an image block division module, an image matching module and an image superposition module; the image decomposition module is used for decomposing a first image and a second image respectively to obtain a first image output result and a second image output result; the image block division module is used for dividing the first image output result into a plurality of first image blocks according to a preset block division manner, the preset block division manner being determined according to the first image output result; the image matching module is used for determining a plurality of second image blocks matched with the plurality of first image blocks from the second image output result based on the plurality of first image blocks, and determining a displacement amount of at least one of the second image blocks relative to the matched first image block in the plurality of second image blocks based on the plurality of first image blocks and the plurality of second image blocks; the image superposition module is used for shifting the plurality of second image blocks based on the displacement amount, and superimposing and denoising the shifted second image and the first image.

[0006] One of the embodiments of the present specification provides a computer readable storage medium, the storage medium stores computer instructions, when the computer reads the computer instructions in the storage medium, the computer executes the image denoising method. BRIEF DESCRIPTION OF DRAWINGS

[0007] The present specification will be further described in the manner of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same numbers refer to the same structures, wherein:

[0008] Figure 1 is a schematic diagram of an application scenario of an image denoising system according to some embodiments of the present specification;

[0009] Figure 2 is a schematic diagram of an image denoising system according to some embodiments of the present specification;

[0010] Figure 3 is an exemplary flowchart of an image denoising method according to some embodiments of the present specification;

[0011] Figure 4 is a schematic diagram of an image denoising method according to some embodiments of the present specification;

[0012] Figure 5 is a schematic diagram of an image denoising method according to some embodiments of the present specification. DETAILED DESCRIPTION

[0013] In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the drawings required to be used in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some examples or embodiments of the present specification, and for those skilled in the art, the present specification can also be applied to other similar scenarios without creative labor. Unless it is obvious from the language environment or otherwise stated, the same reference numbers in the drawings represent the same structure or operation.

[0014] It should be understood that the "system", "device", "unit" and / or "module" used herein is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0015] As used in the description and the claims, the indefinite articles "a", "an", and / or "the" are not intended to mean only one unless otherwise indicated by the context. Generally, the terms "including," "includes", "having", "has", "containing", "contains", or the like are intended to be inclusive (i.e., to mean "comprising") and not to exclude or be limited to whatever other steps, elements, or elements are explicitly stated. In general, the terms "including", "includes", "containing", "contains", "having", "has", or the like are not intended to be limiting.

[0016] Flow diagrams are used in this specification to illustrate the operations according to embodiments of the present specification. It should be understood that the operations in the figures do not necessarily have to be performed in the order shown. Rather, the operations can be performed in different order or simultaneously. Other operations can also be added or removed from the processes.

[0017] In some application scenarios, the image denoising system can include a processing device, a medical imaging device, and the image denoising system can implement motion compensation for the image acquired by the medical imaging device by implementing the methods and / or processes disclosed in the present specification through the processing device, etc., so as to remove the motion artifacts in the multi-frame superposition process, effectively reduce the noise in the image, improve the image quality, and improve the diagnosis effect.

[0018] Figure 1 is a schematic diagram of an application scenario of an image denoising system according to some embodiments of the present specification.

[0019] As Figure 1 shown, in some embodiments, the system 100 can include a medical imaging device 110, a processing device 120, a storage device 130, a terminal 140, and a network 150.

[0020] The medical imaging device 110 refers to a device that reproduces the structure inside the human body as an image using different media in the medical field. In some embodiments, the medical imaging device 110 can be any medical device that images or treats a specified body part of a patient, such as a digital subtraction angiography (DSA) device, a mobile C-arm device, a gastrointestinal machine, etc. The medical imaging device 110 provided above is for illustrative purposes only and is not intended to limit the scope thereof. In some embodiments, the medical imaging device 110 can acquire a plurality of scan images, such as a plurality of consecutive images, etc., and send them to the processing device 120. The medical imaging device 110 can receive instructions, etc. sent by the physician through the terminal 140 and perform relevant operations, such as imaging, etc. according to the instructions. In some embodiments, the medical imaging device 110 can exchange data and / or information with other components (e.g., the processing device 120, the storage device 130, the terminal 140) in the system 100 through the network 150. In some embodiments, the medical imaging device 110 can be directly connected to other components in the system 100. In some embodiments, one or more components (e.g., the processing device 120, the storage device 130) in the system 100 can be included in the medical imaging device 110.

[0021] The processing device 120 can process data and / or information obtained from other devices or system components, perform the image denoising method shown in some embodiments of the present specification based on the data, information and / or processing results, and complete one or more functions described in some embodiments of the present specification. For example, the processing device 120 can eliminate motion artifacts by aligning the same structure in the images by shifting the image blocks based on a plurality of scan images of the medical imaging device 110. For another example, the processing device 120 can superimpose a plurality of images to obtain a denoised image. In some embodiments, the processing device 120 can send the data processed, such as the block manner of the image, the block result, the image after the shift processing, the denoised image, etc., to the storage device 130 for storage. In some embodiments, the processing device 120 can obtain pre-stored data and / or information, such as a formula for calculating the similarity of the image blocks, a wavelet packet decomposition algorithm, etc., from the storage device 130 to perform the image denoising method shown in some embodiments of the present specification, such as determining the matching image blocks based on the similarity, etc.

[0022] In some embodiments, processing device 120 can include one or more sub-processing devices (e.g., single-core processing devices or multi-core multi-threaded processing devices). By way of example only, processing device 120 can include a central processing unit (CPU), an application-specific integrated circuit (ASIC), an application-specific instruction-set processor (ASIP), a graphics processing unit (GPU), a physics processing unit (PPU), a digital signal processor (DSP), a field-programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set computer (RISC), a microprocessor, or the like, or any combination thereof.

[0023] Storage device 130 can store data or information generated by other devices. In some embodiments, storage device 130 can store data and / or information acquired by medical imaging device 110, e.g., scan images, etc. In some embodiments, storage device 130 can store data and / or information processed by processing device 120, e.g., shifted images, de-noised images, etc. Storage device 130 can include one or more storage components, each of which can be a standalone device or a part of other devices. Storage device can be local or implemented through cloud.

[0024] Terminal 140 can control the operation of medical imaging device 110. A physician can issue operation instructions to medical imaging device 110 through terminal 140 to make medical imaging device 110 complete specified operations, e.g., irradiate a specified body part of a patient for imaging. In some embodiments, terminal 140 can instruct processing device 120 to perform the image de-noising method as shown in some embodiments of the present specification. In some embodiments, terminal 140 can receive the de-noised image processed by multi-frame superposition from processing device 120. In some embodiments, terminal 140 can be one or any combination of mobile device 140-1, tablet computer 140-2, laptop computer 140-3, desktop computer, and other devices with input and / or output functions.

[0025] Network 150 can connect the components of the system and / or connect the system with external resource parts. Network 150 enables communication between the components and / or between the system and other parts outside the system, facilitating exchange of data and / or information. In some embodiments, one or more components in system 100 (e.g., medical imaging device 110, processing device 120, storage device 130, terminal 140) can send data and / or information to other components through network 150. In some embodiments, network 150 can be any one or more of wired networks or wireless networks.

[0026] It should be noted that the above description is provided for illustrative purposes only and is not intended to limit the scope of this specification. Various changes and modifications can be made by those skilled in the art based on the content of this specification. Features, structures, methods, and other features of the exemplary embodiments described herein can be combined in various ways to obtain other and / or alternative exemplary embodiments. For example, the processing device 120 may be based on a cloud computing platform, such as a public cloud, private cloud, community cloud, and hybrid cloud. However, these changes and modifications will not depart from the scope of this specification.

[0027] Figure 2 This is a schematic diagram of an image noise reduction system according to some embodiments of this specification.

[0028] like Figure 2 As shown, in some embodiments, system 200 may include image decomposition module 210, image segmentation module 220, image matching module 230 and image overlay module 240.

[0029] In some embodiments, the image decomposition module 210 can be used to decompose the first image and the second image respectively to obtain the first image output result and the second image output result.

[0030] In some embodiments, the decomposition method of the first image and the second image may include one of wavelet packet decomposition, pyramid decomposition, etc.

[0031] In some embodiments, the first image and the second image may include X-ray images, etc.

[0032] In some embodiments, the number of decomposed layers can be determined based on high-frequency information in the first and second images or preset layer values, etc.

[0033] In some embodiments, the image segmentation module 220 can be used to divide the first image output result into multiple first image blocks according to a preset segmentation method, wherein the preset segmentation method can be determined according to the first image output result.

[0034] In some embodiments, the image segmentation module 220 may also obtain the structural information of each layer in the first image output result; and determine a preset segmentation method based on the structural information of each layer in the first image output result.

[0035] In some embodiments, for a region in the first image that includes a linear object, the image segmentation module 220 can segment the image based on the linear direction of the linear object, wherein the linear direction of the linear object conforms to a preset relationship with the first image block that includes the linear object, for example, the linear direction has the largest overlap rate with the diagonal of the image block.

[0036] In some embodiments, the linear object can include at least one of a blood vessel, a guide wire, a catheter, and the like.

[0037] In some embodiments, the image matching module 230 can be configured to determine, based on the plurality of first image blocks, a plurality of second image blocks matched with the plurality of first image blocks from the second image output result, and determine, based on the plurality of first image blocks and the plurality of second image blocks, a displacement amount of at least one second image block in the plurality of second image blocks relative to the matched first image block.

[0038] In some embodiments, the image matching module 230 can determine whether a first image block is matched with a second image block according to a similarity between the first image block and the second image block.

[0039] In some embodiments, the image matching module 230 can determine, layer by layer from back to front according to the number of decomposed layers, a plurality of second image blocks matched with the plurality of first image blocks from the second image output result.

[0040] In some embodiments, the image matching module 230 can determine, based on the displacement amount calculated in the first image output result and the second image output result, an initial value of the displacement amount of at least one second image block in the remaining layers, and determine the displacement amount of the second image block matched with the first image block in the remaining layers based on the initial value.

[0041] In some embodiments, the image superimposition module 240 can be configured to shift the plurality of second image blocks based on the displacement amount, and superimpose and denoise the shifted second image and the first image.

[0042] In some embodiments, the image superimposition module 240 can shift the plurality of second image blocks based on the displacement amount to obtain a shifted second image as a third image, and superimpose and denoise the third image and the first image.

[0043] Figure 3 is an exemplary flowchart of an image denoising method according to some embodiments of the present specification.

[0044] As shown in Figure 3 In some embodiments, the flow 300 can include the following steps. In some embodiments, the flow 300 can be performed by the processing device 120.

[0045] Step 310, decompose the first image and the second image respectively to obtain a first image output result and a second image output result. In some embodiments, step 310 can be performed by the image decomposition module 210.

[0046] The first image and the second image are images acquired based on one scan of the same target object (e.g., a patient), such as images acquired in succession, superimposed images, and the like. In some embodiments, the first image and the second image can be images acquired in succession or with a short interval, such as two consecutive images, two images with one or two frames apart, and the like. In some embodiments, the first image can include a current frame (i.e., the latest frame) image acquired, and the second image can include a previous frame image of the current frame. In some embodiments, the first image can include a current frame (i.e., the latest frame) image acquired, and the second image can include a superimposed image of a plurality of frames acquired before the current frame. In some embodiments, the first image and the second image can be acquired by scanning the target object by a medical imaging device. In some embodiments, the first image and the second image can be acquired by a storage device or other means, which is not limited in the present specification.

[0047] In some embodiments, the first image and the second image can include X-ray images, such as DSA images, and the like. In some embodiments, the first image and the second image can reflect blood vessel information of the target object.

[0048] The first image output result is an image processing result corresponding to the first image, and the second image output result is an image processing result corresponding to the second image. The image output result can be represented by an image or a numerical value, such as a multi-layer image containing low-frequency and / or high-frequency information, a coefficient value, a function value, and the like. In some embodiments, the image output result can include a result of image decomposition (e.g., Wavelet Packet Decomposition, pyramid decomposition, and the like). In some embodiments, the image output result can include a result of Wavelet Packet Decomposition, and can include a plurality of coefficient values, the physical meaning of which is the projection value of the original signal on each wavelet function base, and the time information corresponding to each coefficient value.

[0049] In some embodiments, the first image and the second image can be decomposed respectively to obtain the first image output result and the second image output result.

[0050] In some embodiments, the first image and the second image can be decomposed respectively by one of Wavelet Packet Decomposition, pyramid decomposition, and the like. In some embodiments, the decomposition manner of the first image and the second image can be the same.

[0051] In some embodiments, the pyramid decomposition can include Gaussian pyramid decomposition, Laplacian pyramid decomposition, and the like.

[0052] The wavelet packet decomposition is a further optimization of the wavelet transform. Based on the wavelet transform, in the signal decomposition at each level, in addition to further decomposing the low-frequency subband (low-frequency information of the image), the high-frequency subband (high-frequency information of the image) is also further decomposed. Finally, the optimal signal decomposition path is calculated by minimizing a cost function, and the original signal is decomposed according to the decomposition path. The original signal can be an image signal collected based on a medical imaging device, for example, an original collected image, an image after superimposition of the original collected image, etc.

[0053] The image decomposition (for example, wavelet packet decomposition, pyramid decomposition, etc.) can divide the image into multiple layers of signals, for example, three layers, four layers, five layers, etc. In some embodiments, the first image output result can include images of each layer after the first image decomposition, and the second image output result can include images of each layer after the second image decomposition. In some embodiments, the number of layers of the first image and the second image decomposition can be equal. In some embodiments, the number of layers of decomposition can be determined according to the high-frequency information in the first image and the second image or a preset number of layers. For example, if the frequency of a certain layer of signal after decomposition is lower than a preset threshold, the decomposition is stopped, and the current number of decomposition layers is taken as the number of wavelet packet decomposition layers. For another example, the preset number of layers can be 3 layers, 4 layers, or 5 layers, etc., and the number of layers of decomposition is the same as the preset number of layers.

[0054] In step 320, the first image output result is divided into a plurality of first image blocks according to a preset block mode. In some embodiments, the preset block mode can be determined according to the first image output result. In some embodiments, step 320 can be performed by the image block module 220.

[0055] The block mode refers to the way of dividing the image into image blocks, which can be various modes, for example, according to the richness of the image structure information, according to whether there is a linear object, average block, etc. In some embodiments, the size of the image block can be preset, for example, one of 4*4, 8*8, 16*16, 32*32, etc. or any combination thereof, wherein the size unit of the block is pixel. In some embodiments, the shape of the image block can be various shapes, for example, rectangular, circular, triangular, irregular, etc. The first image block is an image block in the first image. In some embodiments, the first image block can include a rectangular block, etc. In some embodiments, the first image output result can be divided into a plurality of first image blocks according to the preset block mode, wherein the preset block mode can be determined according to the first image output result.

[0056] In some embodiments, each layer after the first image decomposition can be divided into a plurality of first image blocks according to the preset block mode, wherein the number and / or size of the first image blocks contained in each layer can be the same or different.

[0057] In some embodiments, the first image can be directly divided into a plurality of first image blocks according to the preset block mode without the image decomposition step. In some embodiments, the first image without the image decomposition step can be equivalent to only one layer.

[0058] In some embodiments, the structure information of each layer in the first image output result can be obtained, and the preset block mode can be determined based on the structure information of each layer in the first image output result. In some embodiments, the preset block mode of each layer can be determined based on the structure information of each layer in the first image output result, i.e., the block mode of each layer can be determined respectively, and the block mode of each layer can be different. In some embodiments, the block mode of each layer can be the same, i.e., the preset block mode of all layers can be determined based on the structure information of a certain layer in the first image output result.

[0059] In some embodiments, the structure information of the first image can be directly obtained, and the preset block mode can be determined based on the structure information of the first image.

[0060] In some embodiments, a smaller block, e.g., a 4*4 block, can be used in a position with rich structure information, e.g., containing a plurality of structures or structure feature information, etc., and a larger block, e.g., a 32*32 block, can be used in a relatively flat area, e.g., the same structure, little change in structure information, etc. In some embodiments, various methods such as image edge detection algorithm can be used to determine whether the image area is flat. For example, an image area containing more structure edge points is determined as an area with rich structure information, and an image area containing no structure edge points or a small number of structure edge points is determined as a flat area.

[0061] In some embodiments of the present specification, by determining the image block mode based on the image structure information, the stability of block matching can be improved, and the accuracy of block matching can be improved.

[0062] The linear object in the image refers to an object in the image with a linear shape, e.g., blood vessels, guide wires, catheters, etc. in the DSA image. In some embodiments, the preset block mode can be determined according to whether there is a linear object in the first image.

[0063] In some embodiments, various methods such as Hessian matrix based method can be used to detect the linear object in the image.

[0064] In some embodiments, for a region in the first image including a linear object, the region can be partitioned based on a linear direction of the linear object, where the linear direction of the linear object meets a preset relationship with a first image block including the linear object. In some embodiments, the preset relationship can include that a diagonal line of the rectangular block (i.e., the first image block including the linear object) has the maximum coincidence rate with the linear direction of the linear object, i.e., the diagonal line coincides with or is closest to coinciding with the linear direction. In some embodiments, the preset relationship can also include other relationships, such as that the linear direction of the linear object is parallel to a side of the rectangular block.

[0065] In some embodiments of the present specification, by determining the image partitioning manner based on whether there is a linear object in the image, various linear objects such as guide wires and catheters in the image can be well processed, so that the related features of the linear objects can be reflected in the partitioning result, and the accuracy of block matching is improved.

[0066] At step 330, based on the plurality of first image blocks, a plurality of second image blocks matched with the plurality of first image blocks are determined from the second image output result, and based on the plurality of first image blocks and the plurality of second image blocks, a displacement amount of at least one second image block in the plurality of second image blocks relative to the matched first image block is determined. In some embodiments, step 330 can be performed by the image matching module 230.

[0067] The second image block is an image block in the second image. In some embodiments, based on the plurality of first image blocks, a plurality of second image blocks matched with the plurality of first image blocks can be determined from the second image output result.

[0068] In some embodiments, for part or all of the first image blocks, a second image block matched with each of the first image blocks can be determined from the second image output result, thereby obtaining a plurality of second image blocks. For example, for each of all the first image blocks, a second image block matched therewith can be determined. For another example, for each of part of the first image blocks satisfying a preset condition (e.g., rich structural information, etc.), a second image block matched therewith can be determined. In some embodiments, the plurality of determined image blocks can overlap. In some embodiments, the sum of the areas covered by the plurality of determined image blocks can not cover the entire area of the second image.

[0069] In some embodiments, for each layer after wavelet packet decomposition, block matching can be performed based on the first image output result and the second image output result, i.e., for each layer in the first image, a second image block matched with at least one first image block in the layer is determined.

[0070] In some embodiments, the shape and size of the first image block and the second image block matched therewith can be the same.

[0071] In some embodiments, for a first image block, a corresponding layer in the second image output result can be determined according to the layer in which the first image block is located, and a second image block matching the first image block can be determined in the corresponding layer. Specifically, a plurality of second image blocks with the same size and shape as the first image block can be obtained in the corresponding layer of the second image output result, and then a second image block matching the first image block can be determined from the plurality of second image blocks according to a preset rule (e.g., maximum similarity).

[0072] In image decomposition, the layer that is decomposed first is referred to as a front layer, and the layer that is decomposed later is referred to as a back layer. For example, the first layer is decomposed first, the second layer is decomposed second, and so on until the last layer is decomposed. The order from the first layer to the last layer is referred to as from front to back. As the number of layers increases from front to back, the effective information of each layer becomes less and less, and therefore the image of each layer becomes smaller and smaller. In some embodiments, the second image block matching the first image block can be determined in the second image output result layer by layer in an order from back to front according to the number of layers of image decomposition. For example, if the first image and the second image are decomposed into 4 layers in total, all matching image blocks in the 4th layer are determined first, then all matching image blocks in the 3rd layer are determined, and so on until all matching image blocks in the 4th layer are determined.

[0073] In some embodiments, the second image block matching the first image block can be determined in the second image output result layer by layer in other orders, such as from front to back. Alternatively, the second image block matching the first image block can be determined in each layer randomly without setting an order.

[0074] In some embodiments, for a first image block that does not satisfy a preset condition (e.g., the structural information is not rich enough), a second image block at a corresponding position in the second image can be determined as the second image block matching the first image block in a corresponding layer in the second image output result according to the position of the first image block in the first image.

[0075] In some embodiments, whether the first image block matches the second image block can be determined according to the similarity between the first image block and the second image block. For example, the second image block with the maximum similarity can be determined as the second image block matching the first image block.

[0076] In some embodiments, the similarity of an image block can be calculated by using a cross-correlation function, a cross-structure function, a histogram-based similarity metric, or the like. In some embodiments, the similarity of an image block can be obtained by summing the similarity of each pixel point in the image block. In some embodiments, the similarity of an image block can be obtained by other means, such as randomly obtaining a preset number of pixel points in the image and taking the average of the similarity of the pixel points as the value of the similarity of the image block.

[0077] In low-dose X-ray images, the main noise is Poisson noise and quantization noise. In some embodiments, the similarity measure of block matching can use a similarity measure function based on Poisson, quantization noise correlation. In some embodiments, the similarity of two pixels can be obtained by the following formula:

[0078]

[0079] where d represents the similarity of two pixels with gray values of k and l respectively, and the smaller d is, the more similar the two pixels are; q k and q l represent the minimum number of photons required when the gray values reach k and l respectively, which is related to the characteristics of the flat panel detector used to collect the image and can be obtained through experiments; λ is a parameter related to Poisson distribution and can be calculated by the following formula:

[0080] λ = (λ k + λ l ) / 2 (2)

[0081] λ k and λ l in formula (2) can be calculated by the following formula:

[0082]

[0083] where m in λ m represents k or l in formula (1) and (2).

[0084] In some embodiments, after determining a plurality of second image blocks matched with a plurality of first image blocks, the displacement of at least one of the second image blocks relative to the matched first image block can be determined based on the first image blocks and the second image blocks. The displacement refers to the relative offset of the coordinates of the two matched image blocks.

[0085] In some embodiments, the displacement of the matched second image block relative to each of the first image blocks can be determined.

[0086] In some embodiments, the displacement of the matched second image block relative to each of the first image blocks that meet the preset rules (e.g., rich structural information, etc.) can be determined.

[0087] In some embodiments, for each of the first image blocks that do not meet the preset rules (e.g., insufficient structural information, etc.), the displacement of the matched second image block relative to the first image block can be determined as 0, that is, the position of the matched second image block in the second image is the same as the position of the first image block in the first image.

[0088] In some embodiments, the initial value of the displacement of at least one second image block in the remaining layers can be determined based on the calculated displacement in the first image output result and the second image output result, and the second image block in the remaining layers that matches the first image block and the corresponding displacement can be determined based on the initial value. For example, the first image and the second image are decomposed into 4 layers, after the matching image block in the 4th layer is determined, the initial search position (i.e. the initial value of the displacement) of the second image block in the 3rd layer can be determined based on the difference in position, size, shape, etc. of the first image block in the 4th layer and the first image block in the 3rd layer, and the second image block in the 3rd layer that matches the corresponding first image block is searched from the initial search position, thereby reducing the search amount and improving the search efficiency. For example, it is assumed that the determined initial search position coordinate value is (10, 10), and the displacement of the determined matching second image block relative to the initial search position in the 3rd layer is (2, 2), then the displacement of the matching second image block is (12, 12).

[0089] The method of the present embodiment combines the above-mentioned manner of determining the second image block that matches the first image block in the second image output result layer by layer in the order from the back to the front according to the number of decomposition layers of the image, and can further improve the search efficiency of the second image block.

[0090] In some embodiments, in each layer, the search for the second image block that matches the first image block in the layer can start from a preset initial position (e.g. the upper left corner, the lower right corner, etc. of the image).

[0091] In some embodiments, the coordinate difference between a certain pixel point (e.g. a boundary point, etc.) in a certain first image block and the pixel point of the second image block that matches it can be obtained and used as the displacement of the second image block. For example, the lower left corner pixel point of the first image block has a coordinate of (x1, y1) in the first image, and the lower left corner pixel point of the second image block that matches it has a coordinate of (x2, y2) in the second image, then the displacement of the second image block relative to the first image block can be represented as (x2-x1, y2-y1).

[0092] In some embodiments, the coordinate differences of multiple pixel points in the image block can be obtained and used as the displacement of the image block, for example, the coordinate differences of the lower left corner pixel points of the first image block and the second image block are (x2-x1, y2-y1), the coordinate differences of the upper right corner pixel points are (x4-x3, y4-y3), and the coordinate differences of the lower right corner pixel points are (x6-x5, y6-y5), then the displacement of the second image block relative to the first image block can include the coordinate differences of the three points. In some embodiments, the displacement can also be determined in other ways, which are not limited in the present specification.

[0093] In some embodiments of the present specification, the similarity of a block is determined by summing the similarity of the pixels in the block, which is simple and intuitive for block matching, has good practicability, and high accuracy.

[0094] At step 340, the plurality of second image blocks are shifted based on the displacement amounts, and the denoising is performed based on the superposition of the shifted second image and the first image. In some embodiments, step 340 can be performed by the image superposition module 240.

[0095] In some embodiments, each second image block can be shifted based on the displacement amount of the second image block relative to the matched first image block, so as to move the second image block to the position of the first image block, i.e., to shift the second image based on the first image, so as to obtain the shifted second image as the third image. As shown in FIG. 5, the image blocks 510-1, 510-2, …, 510-n in the first image 510 are matched with the image blocks 520-1, 520-2, …, 520-n in the second image 520, respectively, and each image block 520-1, 520-2, …, 520-n in the second image can be shifted based on the respective displacement amount, so as to obtain the shifted second image 520. Figure 5

[0096] In some embodiments, for the first image block for which no matched image block is determined, the displacement amount of the second image block relative to the first image block can be considered as equal to 0, i.e., a second image block with the same shape and size as the first image block in the second image is determined as the shifted second image block, wherein the position of the second image block in the second image is the same as the position of the first image block in the first image.

[0097] In theory, if it is a translational motion, the displacement amounts of the plurality of second image blocks relative to the matched first image block in the second image can be equal, i.e., the motion of the plurality of corresponding positions in the image can be consistent. In some embodiments, the displacement amount of one of the second image blocks in the second image can be taken as the displacement amount of the plurality of second image blocks, and then the second image blocks are shifted based on the displacement amount. For example, the displacement amount of the second image block relative to the first image block can be represented as (x2-x1, y2-y1), and the coordinates of the pixels of all the second image blocks can be subtracted by the displacement amount (x2-x1, y2-y1), so as to obtain the new coordinates as the coordinates of the pixels of the shifted second image block, i.e., the coordinates of the lower left corner pixel of the second image block in the second image are moved from (x2, y2) to (x1, y1).

[0098] ​If the motion includes a rotation, the displacement amount of the plurality of second image blocks relative to the matched first image block can be different. In some embodiments, each second image block can be shifted according to its displacement amount. For example, if the first image block and the second image block have a left bottom pixel coordinate difference of (x2-x1, y2-y1), a right top pixel coordinate difference of (x4-x3, y4-y3), and a right bottom pixel coordinate difference of (x6-x5, y6-y5), and (x2-x1)≠(x4-x3)≠(x6-x5), then the left bottom pixel coordinate of the second image block is shifted from (x2, y2) to (x1, y1), the right top pixel coordinate is shifted from (x4, y4) to (x3, y3), and the right bottom pixel coordinate is shifted from (x6, y6) to (x5, y5).

[0099] In some embodiments, the third image, i.e., the shifted second image, can be superimposed with the first image to obtain a denoised image. As shown in FIG. 5B, the first image 510 can be superimposed with the shifted second image 520 to obtain a denoised first image 530, where the image blocks 530-1, 530-2, …, 530-n contained in the denoised first image 530 are respectively matched with the image blocks 510-1, 510-2, …, 510-n in the first image 510 and the image blocks 520-1, 520-2, …, 520-n in the second image 520, 530-1 is equivalent to the superposition of 510-1 and 520-1, 530-2 is equivalent to the superposition of 510-2 and 520-2, and the remaining image blocks contained in 530 are similar. Figure 5

[0100] In some embodiments, superposition can be performed on each layer of the first image and the second image decomposition respectively, and the layers after superposition are combined to obtain a denoised image.

[0101] In some embodiments, superposition can be directly performed on the original images of the first image and the second image to obtain a denoised image.

[0102] In some embodiments, the image after superposition of the third image and the first image can be further superimposed with a fourth image to obtain a denoised image, where the fourth image can be an image that is time-continuous or has a very short interval with the first image and the second image. In some embodiments, a plurality of continuous images or a plurality of single images with a very short interval can be obtained, and then the shifted images of the other images based on one of the images are obtained, and then the image is superimposed with all the shifted images to obtain a denoised image corresponding to the image.

[0103] ​In some embodiments of the present specification, by shifting the image blocks in the image according to the displacement amount, the structural displacement caused by motion is eliminated, the motion artifact is removed, the influence of motion on the image is well eliminated, thereby greatly improving the effect of superposition denoising, clear and explicit image structure information (for example, blood vessel information, etc.) can be obtained, the quality of the image is improved, and the diagnostic quality is improved.

[0104] Figure 4 is a schematic diagram of an image denoising method according to some embodiments of the present specification.

[0105] In some embodiments, the image can be denoised by the flow 400 shown in the figure. Figure 4 In some embodiments, the flow 400 can be executed by the processing device 120.

[0106] In some embodiments, the first image can include a current frame image 410, and the second image can include a previous frame image 420, wherein the current frame image 410 and the previous frame image 420 are two consecutive frames of images, the current frame image 410 is an image of a current frame, and the previous frame image 420 is an image of a frame before the current frame image 410. For more information about the first image and the second image, see the related description of step 310, which will not be repeated here.

[0107] In some embodiments, the image decomposition module 210 can perform wavelet packet decomposition on the current frame image 410 to obtain a current frame image output result 430, wherein the current frame image output result 430 can include multiple layers of images of the current frame image 410 after wavelet packet decomposition; and perform wavelet packet decomposition on the previous frame image 420 to obtain a previous frame image output result 440, wherein the previous frame image output result 440 can include multiple layers of images of the previous frame image 420 after wavelet packet decomposition. The number of layers of wavelet packet decomposition of the current frame image 410 and the previous frame image 420 can be the same. For more information about how to perform wavelet packet decomposition on the image, see the related description of step 310, which will not be repeated here.

[0108] In some embodiments, the image blocking module 220 can obtain the structure information of each layer in the current frame image output result 430, and then determine the blocking manner 435 based on the structure information, that is, the preset blocking manner. For more information about how to determine the preset blocking manner, see the related description of step 320, which will not be repeated here.

[0109] In some embodiments, the image blocking module 220 can block each layer of the current frame image output result 430 based on the blocking manner 435 to obtain multiple current frame image blocks 450. For more information about how to block the image, see the related description of step 320, which will not be repeated here.

[0110] In some embodiments, the image matching module 230 can determine, based on the plurality of current frame image blocks 450, from the previous frame image output 440, a plurality of previous frame image blocks 460 that match the plurality of current frame image blocks 450 through block matching. More on how to perform block matching can be found in the description of step 330, which will not be repeated here.

[0111] In some embodiments, the image matching module 230 can determine, based on the plurality of current frame image blocks 450 and the plurality of previous frame image blocks 460, a displacement amount 470 of at least one of the plurality of previous frame image blocks 460 relative to the matching current frame image block. More on how to determine the displacement amount can be found in the description of step 330, which will not be repeated here.

[0112] In some embodiments, the image superimposition module 240 can shift the plurality of previous frame image blocks 460 based on the displacement amount 470 to obtain shifted previous frame image 480, and then superimpose and denoise the current frame image 410 and the shifted previous frame image 480 to obtain the denoised image 490. More on how to shift the image blocks and superimpose and denoise can be found in the description of step 340, which will not be repeated here.

[0113] It should be noted that the above description of the processes 300, 400 is merely for example and illustration, and does not limit the scope of the present specification. Those skilled in the art can make various modifications and changes to the processes 300, 400 under the guidance of the present specification. However, these modifications and changes are still within the scope of the present specification. For example, the first image and the second image can be exchanged, i.e., the block division manner is determined based on the second image and the blocks are divided based on the second image, the matching first image blocks in the first image are determined based on the second image blocks, the first image blocks are shifted, etc.

[0114] The beneficial effects that the embodiments of the present specification can bring include but are not limited to: (1) by shifting according to the displacement amount of the matching image blocks in different images, the influence of moving objects is well eliminated, the motion artifacts are removed, the effect of multi-frame superposition denoising is greatly improved, the accuracy and resolution of image structure information (for example, blood vessel information, etc.) are improved, the quality of the image is improved, and the diagnostic quality is improved; (2) by determining the image blocking mode based on image structure information, the stability of block matching is increased, and the quality of block matching is improved; (3) by determining the image blocking mode by judging whether there is a linear object, various linear objects such as guide wires and catheters in the image are well handled, so that the related characteristics of the linear objects can be reflected in the image blocking result, and the accuracy and comprehensiveness of block matching are improved; (4) by summing the similarity of the pixels in the block to determine the similarity of the block, the matching image block is determined, the method is simple and intuitive, has good practicability, and has high accuracy. It should be noted that different embodiments can have different beneficial effects, and in different embodiments, the beneficial effects that can be produced can be any one or a combination of the above, or any other beneficial effects that can be obtained.

[0115] The above has described the basic concepts, and it is obvious that the above detailed disclosure is only as an example and does not constitute a limitation on the present specification. Although it is not explicitly stated here, those skilled in the art can make various modifications, improvements and corrections to the present specification. Such modifications, improvements and corrections are suggested in the present specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of the present specification.

[0116] At the same time, specific words are used in the present specification to describe the embodiments of the present specification. As "one embodiment", "an embodiment", and / or "some embodiments" mean a certain feature, structure or characteristic related to at least one embodiment of the present specification. Therefore, it should be emphasized and noted that the "an embodiment" or "one embodiment" or "one alternative embodiment" mentioned in different places in the present specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of the present specification can be properly combined.

[0117] Furthermore, the order of the processing elements and sequences described in this specification are not intended to be construed as a limitation, unless specifically stated, but are included to provide a complete description of one or more embodiments of the present specification. Regardless of the particular sequence of processing elements and sequences, however, the description herein of a process should be understood to include any and all combinations of one or more elements of a process independently selected from each sequence. For example, although the system components described above can be implemented by hardware devices, they can also be implemented by software solutions, such as installing the described system on an existing server or mobile device.

[0118] Similarly, it is to be noticed that the term "comprising", used in the description, is not intended to exclude other elements or steps. It is to be understood that the description and the examples are intended to be illustrative, but not limiting, of the scope of the present specification. Thus, the scope of the present specification should be given by the appended claims, along with their full scope of equivalents, and not by an restricting interpretation of the description or the examples.

[0119] Some embodiments use numerical values in describing components, quantities of ingredients. It should be understood that such numerical values used in describing embodiments can in some examples be modified by the adjectives "about", "approximately", or "substantially". Unless otherwise stated, "about", "approximately", or "substantially" indicate that the described numerical value allows for a ±20% variation. Accordingly, numerical values used in the specification and claims are approximations that can vary depending upon the desired properties sought to be obtained by the individual embodiment. In some embodiments, numerical values are determined without considering significant digits before the decimal point and are rounded, according to the normal rules of rounding. Although the numerical ranges and parameters setting forth the broad scope of the embodiments of the specification are approximations, the numerical values set forth in the specific examples are reported as precisely as practicable. The numerical values set forth in the specific examples are provided to be as precise as practicable.

[0120] Each patent, patent application, publication, document, article, book, specification, and other material cited in this specification is hereby incorporated by reference in its entirety for all purposes to the same extent as if each individual publication, document, article, book, specification or other material were specifically and individually indicated to be incorporated by reference in its entirety for its cited and recited purposes. Except in the Examples, or where otherwise explicitly indicated, all numerical quantities in this description are understood to be modified by the word "about". Unless otherwise indicated, the use of the term "or" in this specification is understood to be inclusive, i.e., the term "A or B" means "A or B or both". As used herein, the term "about" means ±20% of the indicated value. As used herein, the term "and / or" means "and" or "or", i.e., "A and / or B" means "A and B", "A or B", "A and", "A or", "B and", "B or", "A and B" or "A or B".

[0121] Finally, it should be understood that the embodiments described herein are only given by way of example and that other modifications can occur to persons skilled in the art. Therefore, the scope of the present description is not intended to be limited to the embodiments described herein but is only limited by the claims that follow.

Claims

1. A method for image denoising, comprising: decomposing a first image and a second image to obtain a first image output and a second image output, respectively; dividing the first image output into a plurality of first image blocks according to a preset block division manner, the preset block division manner being determined according to the first image output, comprising: obtaining structure information of each layer in the first image output; determining the preset block division manner based on the structure information of each layer in the first image output, comprising: determining the preset block division manner of each layer in the first image output based on the structure information of the layer, respectively, the preset block division manner of each layer in the first image output being different; determining a plurality of second image blocks matched with the plurality of first image blocks from the second image output based on the plurality of first image blocks, and determining a displacement amount of at least one of the second image blocks relative to the matched first image block based on the plurality of first image blocks and the plurality of second image blocks; performing a shift on the plurality of second image blocks based on the displacement amount, and performing a superposition denoising on the shifted second image and the first image; dividing the first image output into a plurality of first image blocks, comprising: for a region in the first image including a linear object, dividing the region based on a linear direction of the linear object, the linear direction of the linear object and the first image block including the linear object satisfying a preset relationship; the preset relationship comprising that a diagonal line of the first image block including the linear object and the linear direction of the linear object have a maximum coincidence rate; the linear object comprising at least one of a blood vessel, a guide wire and a catheter.

2. The method of claim 1, wherein a number of layers of the decomposing is determined according to high frequency information in the first image and the second image or a preset layer number value, and / or the decomposing comprises one of wavelet packet decomposition and pyramid decomposition.

3. The method of claim 1, wherein the first image and the second image comprise X-ray images.

4. The method of claim 1, wherein the determining the plurality of second image blocks matched with the plurality of first image blocks from the second image output based on the plurality of first image blocks comprises: determining whether the first image block is matched with the second image block according to similarity of the first image block and the second image block.

5. The method of claim 1, wherein the determining the plurality of second image blocks matched with the plurality of first image blocks from the second image output based on the plurality of first image blocks comprises: determining the plurality of second image blocks matched with the plurality of first image blocks from the second image output layer by layer in a sequence from back to front according to the number of layers of the decomposing.

6. The method of claim 1, wherein the determining the displacement amount of at least one of the second image blocks relative to the matched first image block based on the plurality of first image blocks and the plurality of second image blocks comprises: determine an initial value of the displacement amount of at least one second image block in a remaining layer based on the displacement amount calculated in the first image output result and the second image output result, and determine the displacement amount of the second image block matching the first image block in the remaining layer based on the initial value. 7.An image denoising system, comprising an image decomposition module, an image block module, an image matching module and an image superposition module; the image decomposition module is configured to decompose a first image and a second image respectively to obtain a first image output result and a second image output result; the image block module is configured to divide the first image output result into a plurality of first image blocks according to a preset block mode, wherein the preset block mode is determined according to the first image output result, comprising: obtaining structure information of each layer in the first image output result; determining the preset block mode based on the structure information of each layer in the first image output result, comprising: determining the preset block mode of each layer based on the structure information of each layer in the first image output result respectively, wherein the preset block mode of each layer in the first image output result is different; the image matching module is configured to determine a plurality of second image blocks matching the plurality of first image blocks from the second image output result based on the plurality of first image blocks, and determine a displacement amount of at least one second image block relative to the matching first image block based on the plurality of first image blocks and the plurality of second image blocks; the image superposition module is configured to shift the plurality of second image blocks based on the displacement amount, and superimpose and denoise the shifted second image and the first image; dividing the first image output result into a plurality of first image blocks comprises: for a region including a linear object in the first image, dividing the region based on a linear direction of the linear object, wherein the linear direction of the linear object and the first image block including the linear object meet a preset relationship; the preset relationship comprises that a diagonal line of the first image block including the linear object and the linear direction of the linear object have the maximum coincidence rate; the linear object comprises at least one of a blood vessel, a guide wire and a catheter. 8.A computer readable storage medium, wherein the storage medium stores computer instructions, and when the computer reads the computer instructions in the storage medium, the computer executes the method according to any one of claims 1 to 6.

Images