Method and system for screening typical region of oversized image

Abstract

The invention provides an oversized image typical region screening method and system. The method comprises the following steps: acquiring a pathological full-view image; inputting the pathological full-view image into a configurable stratified sampling network, wherein the configurable stratified sampling network is used for stratified sampling of the pathological full-view image to screen out a typical region; the configurable layered sampling network comprises cascaded multi-stage sampling modules; and the j -level sampling module performs typical region screening on the low-resolution image of the sub-image block output by the previous-level sampling module and maps the low-resolution image to the pathological full-view image with the original resolution to obtain a plurality of pathological image blocks with the original resolution and the same region range as the currently screened region. The method can effectively process pathological full-view images with huge resolution and screen out typical areas.

Description

Technical Field

[0001] The present invention relates to the technical field of computer image processing, and in particular to a method and system for screening typical areas of a super-large image. Background Art

[0002] Ovarian cancer is one of the most deadly cancers of the female reproductive system. High-grade serous adenocarcinomas (HGSA) account for 75% of ovarian cancer cases and 80-90% of deaths, and have the worst prognosis among all ovarian cancer subtypes. Even with the same treatment, there are still significant differences in the prognosis of patients with advanced ovarian HGSA. A small number of patients will not relapse in the short term after treatment and have a longer survival period, while most patients (>75%) will relapse within two years.

[0003] The prognosis prediction of ovarian HGSA has important clinical value. Therefore, with the advancement of science and technology in recent years, technicians in this field have added artificial intelligence algorithms to perform prognosis prediction.

[0004] The application of artificial intelligence algorithms in ovarian HGSA analysis is concentrated in the field of bioinformatics mining. In recent years, deep learning methods that can automatically learn from images and quickly obtain excellent feature expressions have achieved great success in the field of computer vision. At the same time, with the advancement of medical imaging and data storage technology, deep learning has been widely used in the field of computer-aided diagnosis. However, even though deep learning has shown great potential, it faces great challenges when applied to pathological image analysis.

[0005] Since the resolution of pathological full-field images is too large, one difficulty in using computer-aided analysis in this field is that tissue pathology full-field images with hundreds of millions or even billions of pixels cannot be directly input into existing deep learning networks for training and reasoning. Summary of the invention

[0006] In order to overcome the defects in the above-mentioned prior art, the object of the present invention is to provide a method and system for screening typical areas of a super-large image.

[0007] In order to achieve the above-mentioned object of the present invention, the present invention provides a method for screening typical regions of a super-large image, comprising the following steps:

[0008] Acquire full-field pathology images;

[0009] The pathological full-field image is input into a configurable hierarchical sampling network, and the configurable hierarchical sampling network is used to perform hierarchical sampling on the pathological full-field image to screen out typical areas.

[0010] The configurable hierarchical sampling network includes cascaded multi-stage sampling modules;

[0011] The first-level sampling module performs typical area screening on the low-resolution image of the pathological full-field image and maps it to the pathological full-field image of the original resolution, thereby obtaining a plurality of pathological image blocks of the original resolution having the same area range as the screened area, namely, the first-level sub-image blocks;

[0012] The j-th level sampling module performs typical area screening on the low-resolution image of the sub-image block output by the sampling module of the previous level and maps it to the pathological full-field image of the original resolution, and obtains multiple pathological image blocks of the original resolution with the same area range as the currently screened area, namely, the j-th level sub-image block, where j is an integer greater than 1.

[0013] This method can effectively process huge resolution full-field pathology images and screen out typical areas. The configurable hierarchical sampling network emphasizes the ability to process a small number of full-resolution image blocks of the input image, which can significantly reduce the peak GPU memory usage and has higher sampling accuracy. For ultra-large images, such as mega- to gigapixel images, the number of selected image blocks is much smaller than the size of the sample space.

[0014] In an optional solution of the typical region screening method for a super large image, each level of the sampling module includes a feature extraction unit, an attention weighting unit and a classification unit;

[0015] The feature extraction unit extracts features from the pathological full-field image or the sub-image block output by the previous sampling module, and the classification unit classifies the features extracted by the feature extraction unit based on the weights of the attention weighting unit to obtain the sub-image block.

[0016] This optional solution achieves the effect of directly using deep neural networks to screen typical feature areas from ultra-large images with hundreds of millions or even billions of pixels (for example: tissue pathology full-field images).

[0017] In an optional solution of the typical region screening method for a very large image, the function of the j-th level sampling module is:

[0018] in, is the classification function of the j-th level sampling module, is the feature extraction function of the j-th level sampling module, is the attention function of the j-th level sampling module, c (j) is the input image x of the j-th level sampling module j The ratio is S j The low-resolution sampling of ∈(0,1) corresponds to the coordinate mapping of the pathological full-field image, and C is the coordinate mapping of all c (j) The set of components;

[0019] The j-1th level sampling function is the input image x to the j-1th level sampling module j-1 The ratio is S j-1 ∈(0,1) low-resolution sampling to obtain the coordinate c (j-1) And c (j-1) Mapped to the j-1th level sampling module input image x j-1 middle, is the j-th level sampling function, the input image x of the j-th level sampling module j for

[0020] This optional solution adopts a multi-layer step-by-step sampling mode, so that the sampling module of each layer only processes the area output by the previous layer, without processing other areas; through the coordinate c (j-1) And the scaling ratio can ensure that the output of the sampling module at each layer can be mapped back to the original image (pathological full-field image).

[0021] In an optional solution of the typical region screening method for the super large image, after Monte Carlo approximation, the function of the j-th level sampling module is: Where M refers to the number of sub-image blocks output by each sampling module, c (j-1) ∈C.

[0022] This alternative can effectively avoid the need to j ,c) calculate features in all image blocks to improve the operation speed.

[0023] In an optional solution of the typical region screening method for a very large image, the loss function of the configurable hierarchical sampling network is:

[0024] in, is the weighted sum of the cross entropy of each level of sampling modules, is the cross entropy of the j-th level sampling module, α (j) is the weight of the j-th level sampling module,

[0025] L RZ =max{||CMLS-SAP1-AM(·)||2,||CMLS-SAP2-AM(·)||2,...,||CMLS-SAP N -AM(·)||2},||CMLS-SAP j -AM(·)||2 is the second-order norm of the network weights of the attention unit of the jth sampling module, β is the weight coefficient of the regularization term, and N is the total number of sampling modules.

[0026] In this optional solution, a regularization term is introduced into the loss function, which allows the model to keep the model simple while minimizing the training error, while improving the generalization performance of the model and preventing overfitting.

[0027] In an alternative solution of the typical region screening method for a very large image, the weight ω1 (j) is the area ratio or quantity ratio of the output image blocks of all levels of sampling modules to the output image blocks of the jth level, ω2 (j) It is the difference between the ratio of the sum of the gradient values ​​of the image block with the smallest sum of gradient information and the image block with the largest sum of gradient information among all the image blocks output by the j-th level sampling module and 1,

[0028] Among them, M (j) is the total number of image blocks output by the j-th level sampling module,

[0029] Among them, tile (j) is the gradient information in the image block output by the j-th sampling module, and m is the sequence number of the image block output by the j-th sampling module.

[0030] The smaller it is, the larger the weight is given to it; the ratio of the number of image blocks output by the l-th level sampling module to the sum of the number of image blocks output by all sampling modules can be directly used to achieve fast approximate calculation. The larger the value is, the higher the information consistency of the output sampled image block is, and a larger weight is given to this part of the output.

[0031] The present invention also proposes a system for screening typical areas of very large images, including an image receiving module, a processing module and a storage module. The image receiving module receives images for training or evaluation, and sends the received images to the processing module. The processing module is communicatively connected to the storage module. The storage module is used to store at least one executable instruction. The executable instruction enables the processing module to perform operations corresponding to the above-mentioned method for screening typical areas of very large images according to the image it receives.

[0032] The beneficial effects of the present invention are:

[0033] The present invention can effectively process the pathological full-field image, use the low-resolution view to indicate the key information area in the reduced view for the purpose of prognostic analysis, and then extract key features in the high-resolution view (pathological full-field image) based on the key information area, without the need to obtain the entire pathological full-field image.

[0034] 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

[0035] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0036] Figure 1 It is a schematic diagram of the structure of a configurable hierarchical sampling network;

[0037] Figure 2 It is a schematic diagram of the structure of the j-th level sampling module in the configurable hierarchical sampling network. DETAILED DESCRIPTION

[0038] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0039] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0040] Embodiment 1

[0041] The present invention provides a method for screening typical regions of a large image, which specifically comprises the following steps:

[0042] Acquire full-field pathology images (reaching billions or even billions of pixels, hence the name ultra-large images).

[0043] In this embodiment, ovarian HGSA patient data is collected as samples. Each patient data includes a full-field pathological image of the primary lesion of ovarian tumor tissue.

[0044] The pathological full-field image is input into a configurable hierarchical sampling network CMLS that has been pre-trained according to screening requirements. The configurable hierarchical sampling network CMLS is used to perform hierarchical sampling on the pathological full-field image to screen out typical areas.

[0045] The configurable layered sampling network CMLS includes cascaded multi-stage sampling modules. The structure of the configurable layered sampling network CMLS in this embodiment is as follows: Figure 1As shown, it includes the first-level sampling module CMLS-SAP1 (·), the second-level sampling module CMLS-SAP2 (·), and the N-level sampling module CMLS-SAP n The number N of the sampling modules is configurable. In this embodiment, the value range of N is 1-5, preferably 2 or 3.

[0046] The first-level sampling module CMLS-SAP1(·) performs typical area screening on the low-resolution image of the pathological full-field image and maps it to the pathological full-field image of the original resolution, and obtains TN1 first-level sub-image blocks of the original resolution with the same area range as the screened area;

[0047] The sampling modules at each level are cascaded in sequence and perform typical area screening on the low-resolution image of the sub-image block output by the sampling module at the previous level and map it to the pathological full-field image at the original resolution, thereby obtaining multiple sub-image blocks at each level with the same area range as the currently screened area.

[0048] That is, the second-level sampling module CMLS-SAP2(·) screens typical regions on the low-resolution image of TN1 first-level sub-image blocks and maps them to the full-field pathological image of the original resolution, and obtains TN2 second-level sub-image blocks of the original resolution with the same regional range as the currently screened region;

[0049] Sequentially, the j-th level sampling module CMLS-SAP j (·) Perform typical area screening on the low-resolution image of the sub-image block output by the previous sampling module and map it to the pathological full-field image of the original resolution to obtain multiple j-th level sub-image blocks of the original resolution with the same area range as the currently screened area, where j is an integer greater than 1.

[0050] Until the Nth level sampling module CMLS-SAP n (·) In TN n-1 The typical regions are screened on the low-resolution image of the (N-1)th level sub-image block and mapped to the full-field pathological image of the original resolution to obtain the TN of the original resolution with the same regional range as the currently screened region. N Nth level sub-image blocks, which contain image detail information for subsequent analysis. The low resolutions collected by different sampling modules in the configurable layered sampling network CMLS can be the same or different, all between 0-1, preferably but not limited to increasing step by step, emphasizing the ability to process a small number of full-resolution image blocks of the input image, which can significantly reduce the peak GPU memory usage and have higher sampling accuracy. For very large images (such as mega- to gigapixel), the number of selected image blocks is much smaller than the sample space || S sampling ||'s size.

[0051] like Figure 2 As shown, each sampling module of the configurable hierarchical sampling network CMLS includes a feature extraction unit CMLS-SAP j -FE(·), attention weighted unit CMLS-SAP j -AM(·) and taxon CMLS-SAP j -CF(·), (1≤j≤N). Feature extraction unit CMLS-SAP j -FE(·) extracts features from the full-field pathology image or the sub-image block output by the previous sampling module, and the classification unit CMLS-SAP j -CF(·) is based on the attention weighted unit CMLS-SAP j The weight of -AM(·) is used to classify the features extracted by the feature extraction unit to obtain the sub-image blocks.

[0052] In this embodiment, the feature extraction unit CMLS-SAP j -FE(·) is composed of trainable parameters A typical example of a deep convolutional neural network for characterization is shown in Table 1.

[0053] Attention Weighted Unit CMLS-SAP j -AM(·) is composed of trainable parameters A smaller deep convolutional neural network represented by is used to obtain the attention weight matrix, as a typical example of which is shown in Table 2.

[0054] Classification unit CMLS-SAP j -CF(·) is composed of trainable parameters A single fully connected layer fc-n representing class .

[0055] Table 1 Feature extraction unit CMLS-SAP j A typical example of -FE(·)

[0056] layer type 1 Conv(3,1,1,8)+ReLU() 2 Conv(3,1,1,16)+ReLU() 3 Conv(3,1,1,32)+ReLU() 4 Conv(3,1,1,16)+ReLU() 5 Conv(3,1,1,8)+ReLU() 6 Conv(3,1,1,16)+ReLU() 7 Conv(3,1,1,32)+ReLU() 8 Conv(3,1,1,1)+ReLU() 9 GlobalAveragePooling2D() 10 SoftMax()

[0057] Table 2 Attention weighted unit CMLS-SAP j A typical example of -AM(·)

[0058] layer type 1 Conv(3,1,1,8)+ReLU() 2 Conv(3,1,1,16)+ReLU() 3 Conv(3,1,1,32)+ReLU() 4 Conv(3,1,1,1)+SoftMax() 5 SoftMax()

[0059] The action function CMLS-SAP1(x) of the sampling module at the first level (i.e. when j=1) of the configurable hierarchical sampling network CMLS is as follows: f ,θ a ,θ c}Parameterized convolutional neural network:

[0060]

[0061] in, is the classification function of the classification unit of the first-level sampling module, is the feature extraction function of the feature extraction unit of the first-level sampling module, and the sampling function From the input image x1∈R H×W Extract the first-level sub-image block of size h1×w1 from the image, which corresponds to the low-resolution view V(x,s1)∈R with a scale of s1∈(0,1) h×w The coordinates c1 = {i, j} in the sampling function, c1 represents the corresponding coordinates of the thumbnail view (the low-resolution view after the original image (pathological full-field image) is reduced) processed by the first-level sampling module and the original image (pathological full-field image), (i, j) is its horizontal and vertical coordinates in the original image, and the sampling function pass Maps coordinate c1 to a position in image x1 and returns a sub-image patch of size h1×w1, where H×W is the resolution of image x1. and h is the height of the thumbnail view of the input level 1 sampling module, w is the width of the reduced view of the input level 1 sampling module, It is a floor operation.

[0062] The attention mechanism is introduced in CMLS-SAP1(x), and the convolutional neural network of the attention mechanism is α=a θ (V(x,s1)), where a θ (·) represents the trainable parameters The smaller deep convolutional neural network represented is used to obtain the attention weight matrix and obtain the attention weighted network:

[0063] Among them, g Θ Generally refers to the classification function in the sampling module, here refers to the classification function in the l-th level sampling module f Θ Generally refers to the feature extraction function in the sampling module, here refers to the feature extraction function in the l-th layer sampling module is the attention matrix, c = (i, j) represents the corresponding coordinates of the thumbnail view processed by the sampling module of this layer and the original image (pathological full-field image), (i, j) represents its horizontal and vertical coordinates in the original image, C is the set of c in the sampling modules of each layer, α c Convolutional neural network with attention mechanism α=a θ (V(x,s1,c)),||S sampling|| is the set of all index pairs of the low-resolution view V(x,s1) of image x1, ||S sampling ||=h×w.

[0064] In order to avoid computing features from all sub-image patches of the low-resolution view V(x,s1), we can modify the formula Perform Monte Carlo approximation to obtain the function of the first-level sampling module: Where C = {(i,j)~a θ (V(x, s1))|i=1,2,...,M} is the index pair M<<||S of the low-resolution view V(x, s1) sampling ||. CMLS-SAP1(x) uses the low-resolution view V(x,s1) of image x1 to calculate the attention weights, obtains M sub-image blocks by sampling, and The average is used to obtain the aggregated features, P refers to the total number of image categories, z n Refers to the features unique to the nth image category.

[0065] Similarly, the action function of the j-th level sampling module of the configurable hierarchical sampling network CMLS is is given by θ={θ f ,θ a ,θ c}Parameterized convolutional neural network.

[0066] in, is the classification function of the classification unit of the j-th level sampling module, is the feature extraction function of the feature extraction unit of the j-th level sampling module, View V(x,s j ,c)∈R u×v is the sampling function Based on scale S j ∈(0,1), c can map the thumbnail view of the layer back to the original image (pathological full-field image). CMLS-SAP j (x) introduces the view V(x,s j ,c) Attention mechanism CMLS-SAP j -AM(·), CMLS-SAP j -AM(·)=V(x,s j ,c)∈R u×v , we get the attention weighted network:

[0067] Where x is the input image, is the attention function CMLS-SAP of the j-th level sampling module j -AM(·), the j-1th level sampling function is the input image x to the j-1th level sampling module j-1 The ratio is S j-1 ∈(0,1) low-resolution sampling to obtain the coordinate c (j-1) And c (j-1) Mapped to the j-1th level sampling module input image x j-1 middle, is the j-th level sampling function, the input image x of the j-th level sampling module j for c (j) is the input image x of the j-th level sampling module j The ratio is S j The low-resolution sampling of ∈(0,1) corresponds to the coordinate mapping of the pathological full-field image, and C is the coordinate mapping of all c (j) The set composed of.

[0068] The low-resolution sampling ratio of the input image for each level of sampling function is set in advance.

[0069] right Perform Monte Carlo approximation to obtain the function of the j-th level sampling module:

[0070]

[0071] Where M refers to the number of sub-image blocks output by each sampling module, c (j-1) ∈C,c (j) ~CMLS-SAP j-1 (V(x,s j-1 ,c (j-1) ))From the distribution V(x,s j-1 ,c (j-1) ) corresponding position extraction, CMLS-SAP j-1 (·) represents the function of the j-1th level sampling module, CMLS-SAP j (x) Use image x j A low-resolution or full-resolution sub-image block from and c (j) The j-level discrete distribution is determined by the j-level low-resolution view V(x,s j ,c (j) ) is obtained. When j = 1, CMLS-SAP j (x) is the selection of typical feature areas on the low-resolution image; subsequently, the typical feature areas are magnified step by step and then sampled until j=N, and then the full-resolution image (pathological full-field image) is performed.

[0072] The loss function of the configurable layered sampling network CMLS is:

[0073]

[0074] in, is the weighted sum of the cross entropy of each level of sampling modules, L RZ is the regularization term, β is the weight coefficient of the regularization term, and the default value is 1.0.

[0075] The weighted sum of the cross entropy of each level of sampling modules Contains the cross entropy of the j-th level sampling module and the weight α of the j-th level sampling module (j) . Weight α (j) It consists of two parts, including the area ratio ω1 of the output / input image of the j-th level sampling module (j) , and the difference between the ratio of the sum of the gradient values ​​of the image block with the smallest sum of gradient information and the image block with the largest sum of gradient information among all the image blocks at the jth level of the output and 1

[0076] Among them, M (j) The total number of image blocks output by the j-th level sampling module,

[0077] Among them, tile (j) is the gradient information in the image block output by the j-th sampling module, and m is the sequence number of the image block output by the j-th sampling module.

[0078] Weight ω1 (j) Represents the area ratio of the output / input image of the j-th sampling module; the smaller the area ratio, the larger the weight is; the ratio of the number of image blocks output by the j-th sampling module to the sum of the number of image blocks output by all sampling modules can be directly used to achieve fast approximate calculation. Weight ω2 (j) Represents the ratio of the minimum value to the maximum value of the gradient information in the output image block of the current j-th layer sampling module. The larger the value, the higher the degree of information consistency of the output sampling image block, and a larger weight is given to this part of the output.

[0079] The purpose of introducing a regularization term into the loss function is to keep the model simple while minimizing the training error, while improving the generalization performance of the model and preventing overfitting.

[0080] L RZ =max{||CMLS-SAP1-AM(·)||2,||CMLS-SAP2-AM(·)||2,...,||CMLS-SAP N -AM(·)||2}, where ||CMLS-SAP j−AM(·)||2 is the second-order norm of the network weights of the attention unit of the jth sampling module.

[0081] The present invention also provides a system for screening typical areas of very large images, comprising an image receiving module, a processing module and a storage module, wherein the image receiving module receives images for training or evaluation, and sends the received images to the processing module, wherein the processing module is communicatively connected to the storage module, and the storage module is used to store at least one executable instruction, wherein the executable instruction enables the processing module to perform operations corresponding to the above-mentioned method for screening typical areas of very large images according to the images it receives.

[0082] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0083] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for screening typical regions of a large image, characterized in that: The following steps are involved: Acquire full-field pathology images; Inputting the pathological full-field image into a configurable hierarchical sampling network, wherein the configurable hierarchical sampling network is used to perform hierarchical sampling on the pathological full-field image to screen out typical areas; The configurable hierarchical sampling network includes cascaded multi-stage sampling modules; The first-level sampling module performs typical area screening on the low-resolution image of the pathological full-field image and maps it to the pathological full-field image of the original resolution, thereby obtaining a plurality of pathological image blocks of the original resolution having the same area range as the screened area, namely, the first-level sub-image blocks; The j-th level sampling module performs typical area screening on the low-resolution image of the sub-image block output by the sampling module of the previous level and maps it to the pathological full-field image of the original resolution, and obtains multiple pathological image blocks of the original resolution with the same area range as the currently screened area, namely, the j-th level sub-image block, where j is an integer greater than 1.

2. The method for screening typical regions of a large image according to claim 1, characterized in that: Each level of the sampling module includes a feature extraction unit, an attention weighting unit and a classification unit; The feature extraction unit extracts features from the pathological full-field image or the sub-image block output by the previous sampling module, and the classification unit classifies the features extracted by the feature extraction unit based on the weights of the attention weighting unit to obtain the sub-image block.

3. The method for screening typical regions of a large image according to claim 1 or 2, characterized in that: The function of the j-th level sampling module is: in, is the classification function of the j-th level sampling module, is the feature extraction function of the j-th level sampling module, is the attention function of the j-th level sampling module, c (j) is the input image x of the j-th level sampling module j The ratio is S j The low-resolution sampling of ∈(0,1) corresponds to the coordinate mapping of the pathological full-field image, and C is the coordinate mapping of all c (j) The set of components; The j-1th level sampling function is the input image x to the j-1th level sampling module j-1 The ratio is s j-1 ∈(0,1) low-resolution sampling to obtain the coordinate c (j-1) And c (j-1) Mapped to the j-1th level sampling module input image x j-1 middle, is the j-th level sampling function, the input image x of the j-th level sampling module j for 4. The method for screening typical regions of a large image according to claim 3, characterized in that: After Monte Carlo approximation, the function of the j-th level sampling module is: Where M refers to the number of sub-image blocks output by each sampling module, c (j-1) ∈C.

5. The method for screening typical regions of a large image according to claim 1, characterized in that: The loss function of the configurable layered sampling network is in, is the weighted sum of the cross entropy of each level of sampling modules, is the cross entropy of the j-th level sampling module, α (j) is the weight of the j-th level sampling module, L RZ =max{||CMLS-SAP1-AM(·)||2,||CMLS-SAP2-AM(·)||2,...,||CMLS-SAP N -AM(·)||2},||CMLS - SAP j -AM(·)||2 is the second-order norm of the network weights of the attention unit of the jth sampling module, β is the weight coefficient of the regularization term, and N is the total number of sampling modules.

6. The method for screening typical regions of a large image according to claim 5, characterized in that: Weight ω1 (j) is the area ratio or quantity ratio of the output image blocks of all levels of sampling modules to the output image blocks of the jth level, ω2 (j) It is the difference between the ratio of the sum of the gradient values ​​of the image block with the smallest sum of gradient information and the image block with the largest sum of gradient information among all the image blocks output by the j-th level sampling module and 1, Among them, M (j) is the total number of image blocks output by the j-th level sampling module, Among them, tile (j) is the gradient information in the image block output by the j-th sampling module, and m is the sequence number of the image block output by the j-th sampling module.

7. A system for screening typical regions of large images, characterized in that: It includes an image receiving module, a processing module and a storage module, the image receiving module receives images for training or evaluation, and sends the received images to the processing module, the processing module is communicatively connected to the storage module, and the storage module is used to store at least one executable instruction, and the executable instruction enables the processing module to perform an operation corresponding to the method for screening typical areas of super-large images as described in any one of claims 1-6 according to the image it receives.

Images