CN116630665A - Mapping data processing method and system applied to image analysis - Google Patents

Abstract

The invention provides a mapping data processing method and system applied to image analysis, and relates to the technical field of artificial intelligence. In the invention, a mapping image to be analyzed is extracted; forming an optimized image analysis network by performing network optimization operation; analyzing analysis possibility parameters of each image pixel in at least two restored drawing images by using an optimized image analysis network; selecting a target restored drawing image from at least two restored drawing images according to analysis possibility parameters of each image pixel in the at least two restored drawing images; and marking the type information of the undetermined image in the target restored drawing image to be marked as the type information of the target image of the drawing image to be analyzed. Based on the above, the reliability of the mapping image analysis can be improved to some extent.

Description

Mapping data processing method and system applied to image analysis

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a mapping data processing method and system applied to image analysis.

Background

In the mapping field, image acquisition is one of the important actions. For the acquired image, based on some requirements, the image needs to be analyzed to determine the kind of the image (such as abnormal kind or deformation kind, kind of the corresponding region, etc.). However, in the prior art, there is a problem that reliability is not high in the process of performing the survey image analysis.

Disclosure of Invention

In view of the above, the present invention is directed to a mapping data processing method and system for image analysis, so as to improve the reliability of the mapping image analysis to a certain extent.

In order to achieve the above purpose, the embodiment of the present invention adopts the following technical scheme:

a mapping data processing method applied to image analysis, comprising:

extracting a mapping image to be analyzed, wherein the mapping image to be analyzed is formed based on target image mapping operation performed on a target area;

forming an optimized image analysis network by performing network optimization operation;

analyzing analysis possibility parameters of each image pixel in at least two restored drawing images by using an optimized image analysis network, wherein the restored drawing images comprise the mapping image to be analyzed and the type information of the image to be determined;

selecting a target reduction drawing image from the at least two reduction drawing images according to analysis possibility parameters of each image pixel in the at least two reduction drawing images;

marking the type information of the undetermined image in the target restored mapping image to be marked as target type information of the mapping image to be analyzed, wherein the target type information is used for reflecting the type of the mapping image to be analyzed.

In some preferred embodiments, in the mapping data processing method applied to image analysis, the step of analyzing, using an optimized image analysis network, analysis probability parameters of each image pixel in at least two restored mapping images includes:

performing data fusion operation on the mapping image to be analyzed and the type information of the image to be determined so as to form one restored mapping image of the at least two restored mapping images; loading the one restored drawing image to the optimized image analysis network, mining the key information characteristic representation of the one restored drawing image by utilizing the optimized image analysis network, and analyzing analysis possibility parameters of each image pixel in the one restored drawing image according to the key information characteristic representation of the one restored drawing image; or alternatively

Loading the mapping image to be analyzed to the optimized image analysis network, and mining out key information characteristic representation of the mapping image to be analyzed by utilizing the optimized image analysis network; and performing an aggregation operation on the key information feature representation of the mapping image to be analyzed and the key information feature representation of the pending image type information by using the optimized image analysis network to form a key information feature representation of one of the at least two restored mapping images; and analyzing analysis possibility parameters of each image pixel in the one restored drawing image according to the key information characteristic representation of the one restored drawing image by utilizing the optimized image analysis network.

In some preferred embodiments, in the mapping data processing method applied to image analysis, the step of forming an optimized image analysis network by performing a network optimization operation includes:

extracting a plurality of primary exemplary mapping images;

analyzing analysis possibility parameters of each image pixel in any one primary exemplary mapping image by utilizing a primary candidate neural network;

performing network optimization operation on the primary candidate neural network according to analysis possibility parameters of each image pixel in the plurality of primary exemplary mapping images to form a primary optimized neural network, wherein the primary optimized neural network comprises a primary key information mining sub-network;

utilizing the primary key information mining sub-network to mine out image key information characteristic representations corresponding to any one primary exemplary mapping image;

performing network optimization operation on the primary key information mining sub-network according to the image key information characteristic representations corresponding to the primary exemplary mapping images so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network;

And excavating a sub-network according to the advanced key information to form a corresponding optimized image analysis network.

In some preferred embodiments, in the mapping data processing method applied to image analysis, the step of performing a network optimization operation on the primary key information mining sub-network according to the image key information feature representations corresponding to each of the plurality of primary exemplary mapping images to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network includes:

analyzing network learning cost parameters corresponding to the primary exemplary mapping images according to the image key information characteristic representations corresponding to the primary exemplary mapping images;

analyzing the network learning cost parameters corresponding to the primary key information mining sub-network according to the network learning cost parameters corresponding to the primary exemplary mapping images;

and carrying out network optimization operation on the primary key information mining sub-network according to the network learning cost parameter corresponding to the primary key information mining sub-network so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network.

In some preferred embodiments, in the above mapping data processing method applied to image analysis, the arbitrary one of the primary exemplary mapping images is an arbitrary one of the initial exemplary mapping images or an exemplary adjustment mapping image corresponding to the arbitrary one of the initial exemplary mapping images, and the exemplary adjustment mapping image corresponding to the arbitrary one of the initial exemplary mapping images belongs to a mapping image formed by adjusting image pixels in the arbitrary one of the initial exemplary mapping images;

the step of analyzing the network learning cost parameter corresponding to each primary exemplary mapping image according to the image key information feature representation corresponding to each primary exemplary mapping image includes:

analyzing network learning cost parameters of any one initial exemplary mapping image according to the image key information characteristic representation of each initial exemplary mapping image and the image key information characteristic representation of the corresponding exemplary adjustment mapping image of each initial exemplary mapping image;

analyzing network learning cost parameters of the exemplary adjustment mapping image corresponding to the any one initial exemplary mapping image according to the image key information characteristic representation of the initial exemplary mapping image and the image key information characteristic representation of the exemplary adjustment mapping image corresponding to the initial exemplary mapping image.

In some preferred embodiments, in the above mapping data processing method applied to image analysis, the step of analyzing, for the arbitrary one initial exemplary mapping image, the network learning cost parameter of the arbitrary one initial exemplary mapping image according to the image key information feature representation of each initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to each initial exemplary mapping image includes:

calculating a first matching degree characterization parameter between the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image;

calculating a second matching degree characterization parameter between the arbitrary initial exemplary mapping image and other initial exemplary mapping images according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the other initial exemplary mapping images, wherein the other initial exemplary mapping images are the initial exemplary mapping images other than the arbitrary initial exemplary mapping image in the respective initial exemplary mapping images;

Calculating a third matching degree characterization parameter between the any one initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image according to the image key information feature representation of the any one initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image;

and calculating the network learning cost parameter of any one initial exemplary mapping image according to the first matching degree characterization parameter, the second matching degree characterization parameter and the third matching degree characterization parameter.

In some preferred embodiments, in the above mapping data processing method applied to image analysis, the step of analyzing the network learning cost parameter of the exemplary adjustment mapping image corresponding to the arbitrary one of the initial exemplary mapping images according to the image key information feature representation of the respective initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the respective initial exemplary mapping image for the arbitrary one of the initial exemplary mapping images includes:

Calculating a first matching degree characterization parameter between the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image;

calculating a fourth matching degree characterization parameter between the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and other initial exemplary mapping images according to the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the image key information feature representation of the other initial exemplary mapping images, wherein the other initial exemplary mapping images belong to the initial exemplary mapping images other than the arbitrary initial exemplary mapping image in the respective initial exemplary mapping images;

calculating a fifth matching degree characterization parameter between the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image according to the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image;

And calculating the network learning cost parameter of the corresponding exemplary adjustment mapping image of any one initial exemplary mapping image according to the first matching degree characterization parameter, the fourth matching degree characterization parameter and the fifth matching degree characterization parameter.

In some preferred embodiments, in the mapping data processing method applied to image analysis, the step of performing a network optimization operation on the primary key information mining sub-network according to the image key information feature representations corresponding to each of the plurality of primary exemplary mapping images to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network includes:

according to the image key information characteristic representation of each primary exemplary mapping image, analyzing image pixel adjustment estimation data of each primary exemplary mapping image, wherein the image pixel adjustment estimation data of the primary exemplary mapping image is an adjusted possibility parameter of each image pixel in the analyzed primary exemplary mapping image;

extracting image pixel adjustment actual data of each primary exemplary mapping image, wherein the image pixel adjustment actual data of the primary exemplary mapping image is actual data of whether each image pixel in the primary exemplary mapping image is adjusted;

And carrying out network optimization operation on the primary key information mining sub-network according to the image pixel adjustment estimation data of each primary exemplary mapping image and the image pixel adjustment actual data of each primary exemplary mapping image so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network.

In some preferred embodiments, in the mapping data processing method applied to image analysis, the step of mining the sub-network according to the advanced key information to form a corresponding optimized image analysis network includes:

extracting a high-level exemplary mapping image and extracting image type identification data corresponding to the high-level exemplary mapping image;

analyzing analysis possibility parameters of each image pixel in at least two fused exemplary mapping images by using a high-level candidate neural network, wherein the fused exemplary mapping images comprise the high-level exemplary mapping images and undetermined image type information, and the high-level candidate neural network comprises the high-level key information mining sub-network;

analyzing a target fusion exemplary mapping image in the at least two fusion exemplary mapping images according to analysis possibility parameters of each image pixel in the at least two fusion exemplary mapping images;

And carrying out network optimization operation on the high-level candidate neural network according to the undetermined image type information in the target fusion exemplary mapping image and the image type identification data of the high-level exemplary mapping image so as to form a corresponding optimized image analysis network.

The embodiment of the invention also provides a mapping data processing system applied to image analysis, which comprises a processor and a memory, wherein the memory is used for storing a computer program, and the processor is used for executing the computer program so as to realize the mapping data processing method applied to image analysis.

The mapping data processing method and system applied to image analysis provided by the embodiment of the invention can be used for extracting the mapping image to be analyzed; forming an optimized image analysis network by performing network optimization operation; analyzing analysis possibility parameters of each image pixel in at least two restored drawing images by using an optimized image analysis network, wherein the restored drawing images comprise a drawing image to be analyzed and type information of the image to be determined; selecting a target restored drawing image from at least two restored drawing images according to analysis possibility parameters of each image pixel in the at least two restored drawing images; and marking the type information of the undetermined image in the target restored drawing image to be marked as the type information of the target image of the drawing image to be analyzed. Based on the foregoing, since the analysis and prediction are directly performed on the restored mapping image including the mapping image to be analyzed and the type information of the image to be determined, rather than only the mapping image to be analyzed, the pertinence of the analysis and prediction is stronger, so that the reliability of the mapping image analysis can be improved to a certain extent, thereby improving the defects in the prior art.

In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

Fig. 1 is a block diagram of a mapping data processing system applied to image analysis according to an embodiment of the present invention.

Fig. 2 is a flowchart illustrating steps involved in a mapping data processing method applied to image analysis according to an embodiment of the present invention.

Fig. 3 is a schematic diagram of each module included in a mapping data processing apparatus for image analysis according to an embodiment of the present invention.

Description of the embodiments

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

As shown in FIG. 1, an embodiment of the present invention provides a mapping data processing system for use in image analysis. Wherein the mapping data processing system applied to image analysis may comprise a memory and a processor.

In detail, the memory and the processor are electrically connected directly or indirectly to realize transmission or interaction of data. For example, electrical connection may be made to each other via one or more communication buses or signal lines. The memory may store at least one software functional module (computer program) that may exist in the form of software or firmware. The processor may be configured to execute the executable computer program stored in the memory, so as to implement the mapping data processing method applied to image analysis provided by the embodiment of the present invention.

It will be appreciated that in some possible embodiments, the Memory may be, but is not limited to, random access Memory (Random Access Memory, RAM), read Only Memory (ROM), programmable Read Only Memory (Programmable Read-Only Memory, PROM), erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), and the like.

It will be appreciated that in some possible embodiments, the processor may be a general purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), a System on Chip (SoC), etc.; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

It will be appreciated that in some possible embodiments, the mapping data processing system applied to image analysis may be a server with data processing capabilities.

With reference to fig. 2, an embodiment of the present invention further provides a mapping data processing method applied to image analysis, which may be applied to the mapping data processing system applied to image analysis. The method steps defined by the flow related to the mapping data processing method applied to image analysis can be realized by the mapping data processing system applied to image analysis.

The specific flow shown in fig. 2 will be described in detail.

Step S110, extracting a mapping image to be analyzed.

In an embodiment of the present invention, the mapping data processing system applied to image analysis may extract a mapping image to be analyzed. The mapping image to be analyzed is formed based on a target image mapping operation performed on a target region. For example, the mapping image to be analyzed belongs to a target image acquisition device, and image acquisition operation is performed on the target area to form the mapping image.

Step S120, forming an optimized image analysis network by performing a network optimization operation.

In the embodiment of the invention, the mapping data processing system applied to image analysis can form an optimized image analysis network by performing network optimization operation. For example, a network optimization operation may be performed on the original image analysis neural network to form an optimized image analysis network.

Step S130, analyzing analysis possibility parameters of each image pixel in at least two restored drawing images by using an optimized image analysis network.

In the embodiment of the invention, the mapping data processing system applied to image analysis can analyze analysis possibility parameters of each image pixel in at least two restored mapping images by using an optimized image analysis network. The reduction drawing image comprises the mapping image to be analyzed and the type information of the image to be analyzed, namely, at least two types of type information of the image to be determined are respectively fused into the mapping image to be analyzed to form at least two reduction drawing images, in addition, the type information of the image to be determined can be text data, so that when fusion is carried out, the text data can be converted into image data, then, the image data and the mapping image to be analyzed are spliced to form the reduction drawing image, characters corresponding to the text data can be arranged in the image data, and other areas can be white areas. The analysis likelihood parameter can be used to reflect a likelihood that a corresponding image pixel will appear (e.g., a likelihood that a pixel value of a pixel point is a certain value).

Step S140, selecting a target restored graphic image from the at least two restored graphic images according to the analysis probability parameters of each image pixel in the at least two restored graphic images.

In an embodiment of the present invention, the mapping data processing system applied to image analysis may select the target restored image from the at least two restored images according to analysis probability parameters of each image pixel in the at least two restored images. For each of the restored graphic images, the product of the analysis likelihood parameters for the individual image pixels in the restored graphic image may be calculated, and then the restored graphic image corresponding to the product having the greatest value may be marked for marking as the target restored graphic image, or selected based on other means.

And step S150, marking the type information of the undetermined image in the target restored drawing image so as to be marked as the type information of the target image of the to-be-analyzed drawing image.

In the embodiment of the invention, the mapping data processing system applied to image analysis can mark the type information of the undetermined image in the target restored mapping image so as to be marked as the type information of the target image of the mapping image to be analyzed. The target image type information is used for reflecting the image type of the mapping image to be analyzed, such as the abnormal type of the region, the characteristic type of the region and the like.

Based on the foregoing, since the analysis and prediction are directly performed on the restored mapping image including the mapping image to be analyzed and the type information of the image to be determined, rather than only the mapping image to be analyzed, the pertinence of the analysis and prediction is stronger, so that the reliability of the mapping image analysis can be improved to a certain extent, thereby improving the defects in the prior art.

It will be appreciated that, in some possible embodiments, step S120 in the foregoing description, that is, the step of forming an optimized image analysis network by performing a network optimization operation, may further include the following implementations:

a plurality of primary exemplary mapping images are extracted, which are indicative exemplary mapping images (i.e., basis for network optimization), and the primary and advanced exemplary mapping images described later are merely for distinguishing from each other without other special meanings;

analyzing analysis possibility parameters of each image pixel in any one primary exemplary mapping image by using a primary candidate neural network, wherein the analysis possibility parameters of each image pixel in any one primary exemplary mapping image are determined by using the primary candidate neural network, and the analysis possibility parameters of each image pixel in any one primary exemplary mapping image are determined based on the image key information characteristic representation of any one primary exemplary mapping image by using the primary candidate neural network, wherein the image key information characteristic representation of any one primary exemplary mapping image comprises the image pixel key information characteristic representation of each image pixel in any one primary exemplary mapping image; in addition, the analysis likelihood parameter for any one image pixel in the primary exemplary mapping image may be a likelihood that the any one image pixel occurs based on at least one image pixel preceding the any one image pixel;

According to analysis possibility parameters of each image pixel in the plurality of primary exemplary mapping images, performing network optimization operation on the primary candidate neural network to form a primary optimized neural network, wherein the primary optimized neural network comprises a primary key information mining sub-network, for each primary exemplary mapping image, the ratio between the analysis possibility parameters of each image pixel in the primary exemplary mapping image and preset parameters can be calculated first, the sum of logarithm operation results of each ratio is calculated to obtain a network optimized cost parameter corresponding to the primary exemplary mapping image, and based on the network optimized cost parameter, performing network optimization operation on the primary candidate neural network to form a primary optimized neural network, namely performing optimization updating on the network parameters of the primary candidate neural network, wherein the preset parameters belong to the network parameters;

the primary key information mining sub-network is utilized to mine out an image key information characteristic representation corresponding to any one primary exemplary mapping image, for example, the primary exemplary mapping image can be mapped to a feature space, then, the mapping result of the feature space can be subjected to linear processing to form the image key information characteristic representation, wherein network parameters mapped to the feature space and subjected to linear processing can be formed in a corresponding network optimization process;

Performing network optimization operation on the primary key information mining sub-network according to the image key information characteristic representations corresponding to the primary exemplary mapping images so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network;

and excavating a sub-network according to the advanced key information to form a corresponding optimized image analysis network.

It will be appreciated that, in some possible embodiments, the step of performing a network optimization operation on the primary key information mining sub-network according to the image key information feature representations corresponding to the respective primary exemplary mapping images to form a high-level key information mining sub-network corresponding to the primary key information mining sub-network may further include the following implementations:

analyzing network learning cost parameters corresponding to the primary exemplary mapping images according to the image key information characteristic representations corresponding to the primary exemplary mapping images;

analyzing the network learning cost parameters corresponding to the primary key information mining sub-network according to the network learning cost parameters corresponding to the primary exemplary mapping images, for example, summing the network learning cost parameters corresponding to the primary exemplary mapping images to obtain the network learning cost parameters corresponding to the primary key information mining sub-network;

And performing network optimization operation on the primary key information mining sub-network according to the network learning cost parameter corresponding to the primary key information mining sub-network to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network, for example, the network parameters of the primary key information mining sub-network can be optimally updated along the direction of reducing the network learning cost parameter corresponding to the primary key information mining sub-network.

It will be appreciated that in some possible embodiments, the arbitrary one of the primary exemplary mapping images is an arbitrary one of the initial exemplary mapping images or an exemplary adjustment mapping image corresponding to the arbitrary one of the initial exemplary mapping images, where the arbitrary one of the initial exemplary mapping images corresponds to a mapping image formed by adjusting image pixels in the arbitrary one of the initial exemplary mapping images (where the number of adjusted image pixels is not limited, such as at least one, where the number of adjusted image pixels is plural, the plural image pixels may be continuous, or discontinuous, and where a difference between an adjusted pixel value of one image pixel and a pixel value before adjustment should be smaller, such as smaller than a preset pixel value, where the preset pixel value may be configured according to an actual requirement), based on this, the analyzing the network learning parameters corresponding to the respective primary exemplary mapping images according to the image key information feature representation corresponding to the plural primary exemplary mapping images may further include:

Analyzing network learning cost parameters of any one initial exemplary mapping image according to the image key information characteristic representation of each initial exemplary mapping image and the image key information characteristic representation of the corresponding exemplary adjustment mapping image of each initial exemplary mapping image;

analyzing network learning cost parameters of the exemplary adjustment mapping image corresponding to the any one initial exemplary mapping image according to the image key information characteristic representation of the initial exemplary mapping image and the image key information characteristic representation of the exemplary adjustment mapping image corresponding to the initial exemplary mapping image.

It will be appreciated that, in some possible embodiments, the step of analyzing the network learning cost parameter of the arbitrary one initial exemplary mapping image according to the image key information feature representation of the respective initial exemplary mapping image and the image key information feature representation of the respective exemplary adjustment mapping image corresponding to the respective initial exemplary mapping image may further include the following implementations:

Calculating a first matching degree characterization parameter, such as cosine similarity and the like, between the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image;

calculating a second matching degree characterization parameter, such as cosine similarity between feature representations, between the arbitrary initial exemplary mapping image and other initial exemplary mapping images, which are initial exemplary mapping images other than the arbitrary initial exemplary mapping image in the respective initial exemplary mapping images, according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the other initial exemplary mapping images;

calculating a third matching degree characterization parameter, such as cosine similarity between feature representations, between the any one initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image according to the image key information feature representation of the any one initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image;

And calculating the network learning cost parameter of any one initial exemplary mapping image according to the first matching degree characterization parameter, the second matching degree characterization parameter and the third matching degree characterization parameter.

It may be appreciated that, in some possible embodiments, the step of calculating the first matching degree characterizing parameter between the arbitrary initial exemplary mapping image and the exemplary adjusted mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjusted mapping image corresponding to the arbitrary initial exemplary mapping image may further include the following implementation matters:

calculating a number product of the image key information feature representation of the any one initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the any one initial exemplary mapping image to output a corresponding feature representation number product, and performing an adjustment operation on the feature representation number product based on a target adjustment parameter to form a first matching degree characterization parameter between the any one initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the any one initial exemplary mapping image, wherein the first matching degree characterization parameter may have a positive correlation corresponding relation with the feature representation number product, and the second matching degree characterization parameter may have a negative correlation corresponding relation with the target adjustment parameter, and the target adjustment parameter may be determined in a manner that: obtaining the maximum number of network optimization stages (which may be preconfigured, i.e. the number of network optimization, the number of stages or the number of rounds, etc.) and obtaining the current number of network optimization stages, calculating the absolute difference between the current number of network optimization stages and the positive correlation value of the maximum number of network optimization stages (the ratio between the positive correlation value and the maximum number of network optimization stages is equal to a fixed value, such as 0.5), calculating the ratio between the absolute difference and the maximum number of network optimization stages, and finally updating (such as adding) the ratio based on a predetermined offset parameter to obtain the target adjustment parameter, wherein the offset parameter may be a value of 0.01, 0.02, 0.03, 0.04, etc.

It may be appreciated that, in some possible embodiments, the step of calculating the network learning cost parameter of the arbitrary initial exemplary mapping image according to the first matching degree characterizing parameter, the second matching degree characterizing parameter and the third matching degree characterizing parameter may further include the following implementations:

calculating an index operation result of the first matching degree representation parameter to obtain a first index operation result, calculating an index operation result of the second matching degree representation parameter to obtain a second index operation result, and calculating an index operation result of the third matching degree representation parameter to obtain a third index operation result;

calculating the sum of the second index calculation result and the third index calculation result to obtain a target index calculation result, calculating a ratio between the first index calculation result and the target index calculation result, performing a logarithm taking operation on the ratio, finally determining a network learning cost parameter of the arbitrary initial exemplary mapping image based on the result of the logarithm taking operation, wherein the network learning cost parameter may be inversely related to the result of the logarithm taking operation, in other embodiments, directly calculating the sum of the second matching degree characterization parameter and the third matching degree characterization parameter, calculating a ratio between the first matching degree characterization parameter and the sum, performing the logarithm taking operation on the ratio, and finally determining the network learning cost parameter of the arbitrary initial exemplary mapping image based on the result of the logarithm taking operation.

It will be appreciated that, in some possible embodiments, the step of analyzing the network learning cost parameter of the exemplary adjustment mapping image corresponding to the any one initial exemplary mapping image according to the image key information feature representation of the respective initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the respective initial exemplary mapping image may further include the following implementations:

calculating a first matching degree characterization parameter between the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image, as described in the foregoing related description;

calculating a fourth matching degree characterization parameter between the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and other initial exemplary mapping images according to the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the image key information feature representation of the other initial exemplary mapping images, wherein the other initial exemplary mapping images belong to initial exemplary mapping images other than the arbitrary initial exemplary mapping image in the respective initial exemplary mapping images, as described in the related description;

Calculating a fifth matching degree characterization parameter between the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image according to the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image, as described in the foregoing related description;

and calculating the network learning cost parameter of the corresponding exemplary adjustment mapping image of any one initial exemplary mapping image according to the first matching degree characterization parameter, the fourth matching degree characterization parameter and the fifth matching degree characterization parameter, as described in the related description.

It will be appreciated that, in some possible embodiments, the step of performing a network optimization operation on the primary key information mining sub-network according to the image key information feature representations corresponding to the respective primary exemplary mapping images to form a high-level key information mining sub-network corresponding to the primary key information mining sub-network may further include the following implementations:

According to the image key information characteristic representation of each primary exemplary mapping image, analyzing image pixel adjustment estimation data of each primary exemplary mapping image, wherein the image pixel adjustment estimation data of the primary exemplary mapping image is an adjusted possibility parameter of each image pixel in the analyzed primary exemplary mapping image;

extracting image pixel adjustment actual data of each primary exemplary mapping image, wherein the image pixel adjustment actual data of the primary exemplary mapping image is actual data of whether each image pixel in the primary exemplary mapping image is adjusted;

according to the image pixel adjustment estimation data of each primary exemplary mapping image and the image pixel adjustment actual data of each primary exemplary mapping image, performing network optimization operation on the primary key information mining sub-network to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network, that is, determining a corresponding network optimization cost parameter based on the difference between the image pixel adjustment estimation data and the image pixel adjustment actual data, and performing network optimization operation on the primary key information mining sub-network based on the network optimization cost parameter to form the advanced key information mining sub-network.

It will be appreciated that, in some possible embodiments, the step of mining the sub-network according to the advanced key information to form a corresponding optimized image analysis network may further include the following implementations:

extracting an advanced exemplary mapping image, and extracting image type identification data corresponding to the advanced exemplary mapping image, wherein the image type identification data can be formed by manual annotation;

analyzing analysis possibility parameters of each image pixel in at least two fused exemplary mapping images by using a high-level candidate neural network, wherein the fused exemplary mapping images comprise the high-level exemplary mapping images and undetermined image type information, and the high-level candidate neural network comprises the high-level key information mining sub-network;

analyzing a target fusion exemplary mapping image in the at least two fusion exemplary mapping images according to analysis possibility parameters of each image pixel in the at least two fusion exemplary mapping images;

according to the target fusion exemplary mapping image to-be-determined image type information and the image type identification data of the advanced exemplary mapping image, performing network optimization operation on the advanced candidate neural network to form a corresponding optimized image analysis network, for example, according to the difference between the to-be-determined image type information and the image type identification data, determining a corresponding network optimization cost parameter, and performing network optimization operation on the advanced candidate neural network based on the network optimization cost parameter to form the corresponding optimized image analysis network.

It will be appreciated that, in some possible embodiments, the step of analyzing the analysis likelihood parameters of each image pixel in at least two fused exemplary mapping images using the advanced candidate neural network may further include the following implementations:

performing a data fusion operation on the advanced exemplary mapping image and the pending image category information, as previously described in relation to form a fused exemplary mapping image of the at least two fused exemplary mapping images;

loading the one fused exemplary mapping image to the advanced candidate neural network, mining a key information feature representation of the one fused exemplary mapping image using the advanced candidate neural network, and analyzing analysis likelihood parameters of each image pixel in the one fused exemplary mapping image according to the key information feature representation of the one fused exemplary mapping image, as described in the foregoing.

It will be appreciated that, in some possible embodiments, the step of analyzing the analysis likelihood parameters of each image pixel in at least two fused exemplary mapping images using the advanced candidate neural network may further include the following implementations:

Loading the advanced exemplary mapping image to load into the advanced candidate neural network, and mining key information characteristic representations of the advanced exemplary mapping image by using the advanced candidate neural network, as described in the related description;

performing an aggregation operation on the key information feature representation of the advanced exemplary mapping image and the key information feature representation of the pending image category information using the advanced candidate neural network to form a key information feature representation of one fused exemplary mapping image of the at least two fused exemplary mapping images, as previously described in relation thereto;

analyzing analysis probability parameters of each image pixel in the fused exemplary mapping image according to the key information characteristic representation of the fused exemplary mapping image by using the advanced candidate neural network, as described in the related description.

It will be appreciated that, in some possible embodiments, the step of analyzing the target fused exemplary mapping image in the at least two fused exemplary mapping images according to the analysis likelihood parameters of each image pixel in the at least two fused exemplary mapping images may further include the following implementations:

According to the analysis possibility parameters of each image pixel in the at least two fused exemplary mapping images, analyzing the image content linking parameters of each fused exemplary mapping image, wherein the image content linking parameters of the fused exemplary mapping images are used for reflecting the linking degree (namely the matching degree between fused image data) between the image contents included in the fused exemplary mapping images;

and screening out a fusion exemplary mapping image with corresponding image content linking parameters matched with preset linking parameters according to the image content linking parameters of each fusion exemplary mapping image, and marking the fusion exemplary mapping image as the target fusion exemplary mapping image, for example, the fusion exemplary mapping image with the maximum image content linking parameters can be determined as the fusion exemplary mapping image with the corresponding image content linking parameters matched with the preset linking parameters, and thus, the fusion exemplary mapping image can be marked as the target fusion exemplary mapping image.

It will be appreciated that, in some possible embodiments, the step of analyzing the image content linking parameters of each of the fused exemplary mapping images according to the analysis likelihood parameters of each of the image pixels in the at least two fused exemplary mapping images may further include the following implementations:

Analyzing the analysis possibility parameters of each fused exemplary mapping image according to the analysis possibility parameters of each image pixel in the at least two fused exemplary mapping images;

and analyzing the image content linking parameters of the fused exemplary mapping images according to the analysis possibility parameters of the fused exemplary mapping images.

It may be appreciated that, in some possible embodiments, the step of analyzing the analysis likelihood parameter of each of the fused exemplary mapping images according to the analysis likelihood parameter of each of the image pixels in the at least two fused exemplary mapping images may further include the following implementation matters:

for each of the fused exemplary mapping images, a fusion operation, such as multiplication, is performed on the analysis likelihood parameters of the individual image pixels in the fused exemplary mapping image to output the analysis likelihood parameters of the fused exemplary mapping image.

It may be appreciated that, in some possible embodiments, the step of analyzing the image content linking parameters of the fused exemplary mapping images according to the analysis likelihood parameters of the fused exemplary mapping images may further include the following implementations:

For each of the fused exemplary mapping images, determining a number of pixels of an image pixel included in the fused exemplary mapping image, and calculating a first negative correlation parameter of an analysis likelihood parameter of the fused exemplary mapping image (e.g., a product between the first negative correlation parameter and the analysis likelihood parameter is equal to a fixed value, such as 0.5, 1, 2), and calculating a second negative correlation parameter of the number of pixels (e.g., a product between the second negative correlation parameter and the number of pixels is equal to a fixed value, such as 0.5, 1, 2), and performing a power operation on the first negative correlation parameter (e.g., performing an operation on the second negative correlation parameter by a power of a corresponding number of times of the first negative correlation parameter) based on the second negative correlation parameter to output an image content engagement parameter of the fused exemplary mapping image, which may have a negative correlation correspondence, such as equal to a fixed value, between the image content engagement parameter and a result of the power operation, for example.

It will be appreciated that, in some possible embodiments, the step S130 in the foregoing description, that is, the step of analyzing, by using the optimized image analysis network, the analysis probability parameters of each image pixel in at least two restored drawing images may further include the following implementations:

Performing data fusion operation on the mapping image to be analyzed and the type information of the image to be determined to form one restored mapping image of the at least two restored mapping images, as described above;

the method comprises the steps of loading the one restored graphic image into the optimized image analysis network, utilizing the optimized image analysis network to mine out the key information feature representation of the one restored graphic image, analyzing analysis possibility parameters of each image pixel in the one restored graphic image according to the key information feature representation of the one restored graphic image, for example, encoding the restored graphic image through the optimized image analysis network to represent the restored graphic image in a vector form to obtain a corresponding key information feature representation, then performing full-connection operation on the key information feature representation to form a key information full-connection feature representation, finally, performing activation operation on the key information full-connection feature representation to form analysis possibility parameters of each image pixel in the one restored graphic image, wherein the activation function can be realized through an activation function, and the activation function can be a function such as a softmax.

It will be appreciated that, in some possible embodiments, the step S130 in the foregoing description, that is, the step of analyzing, by using the optimized image analysis network, the analysis probability parameters of each image pixel in at least two restored drawing images may further include the following implementations:

loading the mapping image to be analyzed to the optimized image analysis network, and mining out key information characteristic representation of the mapping image to be analyzed by utilizing the optimized image analysis network; and performing an aggregation operation, such as a cascade combination operation, on the key information feature representation of the mapping image to be analyzed and the key information feature representation of the type information of the image to be determined by using the optimized image analysis network to form a key information feature representation of one of the at least two restored mapping images, that is, feature mining can be performed on the mapping image to be analyzed and the type information of the image to be determined respectively to form corresponding key information feature representations respectively; and analyzing analysis possibility parameters of each image pixel in the one restored drawing image according to the key information characteristic representation of the one restored drawing image by utilizing the optimized image analysis network, as described in the previous related description.

With reference to fig. 3, an embodiment of the present invention further provides a mapping data processing apparatus applied to image analysis, which is applicable to the mapping data processing system applied to image analysis. Wherein the mapping data processing apparatus applied to image analysis may include:

the system comprises a mapping image extraction module, a mapping image analysis module and a mapping image analysis module, wherein the mapping image extraction module is used for extracting a mapping image to be analyzed, and the mapping image to be analyzed is formed based on target image mapping operation performed on a target area;

the neural network optimization module is used for forming an optimized image analysis network by performing network optimization operation;

the mapping image analysis module is used for analyzing analysis possibility parameters of each image pixel in at least two restored mapping images by utilizing an optimized image analysis network, and the restored mapping images comprise the mapping image to be analyzed and the type information of the image to be determined;

the mapping image determining module is used for selecting a target restored mapping image from the at least two restored mapping images according to analysis possibility parameters of each image pixel in the at least two restored mapping images;

the image type information determining module is used for marking the undetermined image type information in the target restored drawing image so as to be marked as target image type information of the to-be-analyzed drawing image, and the target image type information is used for reflecting the image type of the to-be-analyzed drawing image.

In summary, the mapping data processing method and system applied to image analysis provided by the invention can extract the mapping image to be analyzed first; forming an optimized image analysis network by performing network optimization operation; analyzing analysis possibility parameters of each image pixel in at least two restored drawing images by using an optimized image analysis network, wherein the restored drawing images comprise a drawing image to be analyzed and type information of the image to be determined; selecting a target restored drawing image from at least two restored drawing images according to analysis possibility parameters of each image pixel in the at least two restored drawing images; and marking the type information of the undetermined image in the target restored drawing image to be marked as the type information of the target image of the drawing image to be analyzed. Based on the foregoing, since the analysis and prediction are directly performed on the restored mapping image including the mapping image to be analyzed and the type information of the image to be determined, rather than only the mapping image to be analyzed, the pertinence of the analysis and prediction is stronger, so that the reliability of the mapping image analysis can be improved to a certain extent, thereby improving the defects in the prior art.

The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A mapping data processing method applied to image analysis, comprising:

extracting a mapping image to be analyzed, wherein the mapping image to be analyzed is formed based on target image mapping operation performed on a target area;

forming an optimized image analysis network by performing network optimization operation;

analyzing analysis possibility parameters of each image pixel in at least two restored drawing images by using an optimized image analysis network, wherein the restored drawing images comprise the mapping image to be analyzed and the type information of the image to be determined;

selecting a target reduction drawing image from the at least two reduction drawing images according to analysis possibility parameters of each image pixel in the at least two reduction drawing images;

marking the type information of the undetermined image in the target restored mapping image to be marked as target type information of the mapping image to be analyzed, wherein the target type information is used for reflecting the type of the mapping image to be analyzed.

2. The method of mapping data processing for image analysis according to claim 1, wherein the step of analyzing the analysis probability parameters of each image pixel in the at least two restored mapping images using the optimized image analysis network comprises:

performing data fusion operation on the mapping image to be analyzed and the type information of the image to be determined so as to form one restored mapping image of the at least two restored mapping images; loading the one restored drawing image to the optimized image analysis network, mining the key information characteristic representation of the one restored drawing image by utilizing the optimized image analysis network, and analyzing analysis possibility parameters of each image pixel in the one restored drawing image according to the key information characteristic representation of the one restored drawing image; or alternatively

Loading the mapping image to be analyzed to the optimized image analysis network, and mining out key information characteristic representation of the mapping image to be analyzed by utilizing the optimized image analysis network; and performing an aggregation operation on the key information feature representation of the mapping image to be analyzed and the key information feature representation of the pending image type information by using the optimized image analysis network to form a key information feature representation of one of the at least two restored mapping images; and analyzing analysis possibility parameters of each image pixel in the one restored drawing image according to the key information characteristic representation of the one restored drawing image by utilizing the optimized image analysis network.

3. A mapping data processing method applied to image analysis according to claim 1 or 2, wherein the step of forming an optimized image analysis network by performing a network optimization operation comprises:

extracting a plurality of primary exemplary mapping images;

analyzing analysis possibility parameters of each image pixel in any one primary exemplary mapping image by utilizing a primary candidate neural network;

performing network optimization operation on the primary candidate neural network according to analysis possibility parameters of each image pixel in the plurality of primary exemplary mapping images to form a primary optimized neural network, wherein the primary optimized neural network comprises a primary key information mining sub-network;

utilizing the primary key information mining sub-network to mine out image key information characteristic representations corresponding to any one primary exemplary mapping image;

performing network optimization operation on the primary key information mining sub-network according to the image key information characteristic representations corresponding to the primary exemplary mapping images so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network;

And excavating a sub-network according to the advanced key information to form a corresponding optimized image analysis network.

4. A mapping data processing method applied to image analysis according to claim 3, wherein the step of performing network optimization operation on the primary key information mining sub-network according to the image key information feature representations corresponding to the respective plurality of primary exemplary mapping images to form a high-level key information mining sub-network corresponding to the primary key information mining sub-network comprises:

analyzing network learning cost parameters corresponding to the primary exemplary mapping images according to the image key information characteristic representations corresponding to the primary exemplary mapping images;

analyzing the network learning cost parameters corresponding to the primary key information mining sub-network according to the network learning cost parameters corresponding to the primary exemplary mapping images;

and carrying out network optimization operation on the primary key information mining sub-network according to the network learning cost parameter corresponding to the primary key information mining sub-network so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network.

5. The method of claim 4, wherein the arbitrary one primary exemplary mapping image is an arbitrary one initial exemplary mapping image or an exemplary adjustment mapping image corresponding to the arbitrary one initial exemplary mapping image, the arbitrary one initial exemplary mapping image corresponding to the exemplary adjustment mapping image belonging to a mapping image formed by adjusting image pixels in the arbitrary one initial exemplary mapping image;

the step of analyzing the network learning cost parameter corresponding to each primary exemplary mapping image according to the image key information feature representation corresponding to each primary exemplary mapping image includes:

analyzing network learning cost parameters of any one initial exemplary mapping image according to the image key information characteristic representation of each initial exemplary mapping image and the image key information characteristic representation of the corresponding exemplary adjustment mapping image of each initial exemplary mapping image;

analyzing network learning cost parameters of the exemplary adjustment mapping image corresponding to the any one initial exemplary mapping image according to the image key information characteristic representation of the initial exemplary mapping image and the image key information characteristic representation of the exemplary adjustment mapping image corresponding to the initial exemplary mapping image.

6. The method of claim 5, wherein the step of analyzing the network learning cost parameter of the arbitrary one initial exemplary mapping image based on the image key information feature representation of each initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to each initial exemplary mapping image for the arbitrary one initial exemplary mapping image comprises:

calculating a first matching degree characterization parameter between the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image;

calculating a second matching degree characterization parameter between the arbitrary initial exemplary mapping image and other initial exemplary mapping images according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the other initial exemplary mapping images, wherein the other initial exemplary mapping images are the initial exemplary mapping images other than the arbitrary initial exemplary mapping image in the respective initial exemplary mapping images;

Calculating a third matching degree characterization parameter between the any one initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image according to the image key information feature representation of the any one initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image;

and calculating the network learning cost parameter of any one initial exemplary mapping image according to the first matching degree characterization parameter, the second matching degree characterization parameter and the third matching degree characterization parameter.

7. The method of claim 5, wherein analyzing the network learning cost parameter of the exemplary adjusted survey image corresponding to the arbitrary one of the initial exemplary survey images based on the image key information feature representation of the respective initial exemplary survey image and the image key information feature representation of the exemplary adjusted survey image corresponding to the respective initial exemplary survey image comprises:

Calculating a first matching degree characterization parameter between the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image according to the image key information feature representation of the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image;

calculating a fourth matching degree characterization parameter between the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and other initial exemplary mapping images according to the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the image key information feature representation of the other initial exemplary mapping images, wherein the other initial exemplary mapping images belong to the initial exemplary mapping images other than the arbitrary initial exemplary mapping image in the respective initial exemplary mapping images;

calculating a fifth matching degree characterization parameter between the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image according to the image key information feature representation of the exemplary adjustment mapping image corresponding to the arbitrary initial exemplary mapping image and the image key information feature representation of the exemplary adjustment mapping image corresponding to the other initial exemplary mapping image;

And calculating the network learning cost parameter of the corresponding exemplary adjustment mapping image of any one initial exemplary mapping image according to the first matching degree characterization parameter, the fourth matching degree characterization parameter and the fifth matching degree characterization parameter.

8. A mapping data processing method applied to image analysis according to claim 3, wherein the step of performing network optimization operation on the primary key information mining sub-network according to the image key information feature representations corresponding to the respective plurality of primary exemplary mapping images to form a high-level key information mining sub-network corresponding to the primary key information mining sub-network comprises:

according to the image key information characteristic representation of each primary exemplary mapping image, analyzing image pixel adjustment estimation data of each primary exemplary mapping image, wherein the image pixel adjustment estimation data of the primary exemplary mapping image is an adjusted possibility parameter of each image pixel in the analyzed primary exemplary mapping image;

extracting image pixel adjustment actual data of each primary exemplary mapping image, wherein the image pixel adjustment actual data of the primary exemplary mapping image is actual data of whether each image pixel in the primary exemplary mapping image is adjusted;

And carrying out network optimization operation on the primary key information mining sub-network according to the image pixel adjustment estimation data of each primary exemplary mapping image and the image pixel adjustment actual data of each primary exemplary mapping image so as to form an advanced key information mining sub-network corresponding to the primary key information mining sub-network.

9. A mapping data processing method for use in image analysis according to claim 3, wherein the step of mining sub-networks according to the advanced key information to form a corresponding optimized image analysis network comprises:

extracting a high-level exemplary mapping image and extracting image type identification data corresponding to the high-level exemplary mapping image;

analyzing analysis possibility parameters of each image pixel in at least two fused exemplary mapping images by using a high-level candidate neural network, wherein the fused exemplary mapping images comprise the high-level exemplary mapping images and undetermined image type information, and the high-level candidate neural network comprises the high-level key information mining sub-network;

analyzing a target fusion exemplary mapping image in the at least two fusion exemplary mapping images according to analysis possibility parameters of each image pixel in the at least two fusion exemplary mapping images;

And carrying out network optimization operation on the high-level candidate neural network according to the undetermined image type information in the target fusion exemplary mapping image and the image type identification data of the high-level exemplary mapping image so as to form a corresponding optimized image analysis network.

10. A mapping data processing system for image analysis, comprising a processor and a memory, the memory for storing a computer program, the processor for executing the computer program to implement the method of any of claims 1-9.