CN113409420A

CN113409420A - User-defined map style drawing method, system, storage medium and equipment

Info

Publication number: CN113409420A
Application number: CN202110958001.7A
Authority: CN
Inventors: 李靖; 郑驰
Original assignee: Shenzhen Topevery Technology Co ltd
Current assignee: Shenzhen Topevery Technology Co ltd
Priority date: 2021-08-20
Filing date: 2021-08-20
Publication date: 2021-09-17

Abstract

The invention relates to the technical field of map generation, in particular to a drawing method, a drawing system, a storage medium and drawing equipment for customizing a map style, wherein the method comprises the following steps: training a model of a convolutional neural network based on the image sample set of the picture; applying the trained convolutional neural network classification model to a new data area to classify the ground objects to obtain a grid data set of ground object classification in the picture; vectorizing the obtained raster data set, and obtaining a vector diagram of each element; and editing the vector diagram by inputting map parameters to complete the map making of the custom style so as to output a custom style map. The method and the device solve the problem that the user is difficult to obtain the map with the custom style, solve a plurality of limitations of the map with the custom style, and enable the user to customize the map with the style according to the preference of the user.

Description

User-defined map style drawing method, system, storage medium and equipment

Technical Field

The invention relates to the technical field of map generation, in particular to a drawing method, a drawing system, a drawing storage medium and drawing equipment for customizing a map style.

Background

The use of the software map becomes a practical tool which cannot be lost in daily life of people more and more, and the software map can embody the practicability of the software map particularly for complex urban roads; and the modes of the large software maps are similar, and the form is single.

At present, a custom style map is generally searched by using a conventional method, and the acquisition difficulty is too high. The custom map style is usually customized by means of authority application, payment calling and the like based on the map platforms such as Baidu and Gaode, and the like, and the custom map style cannot be used offline and has too many constraint conditions.

Due to the complexity of the production flow of surveying and mapping results and the dependence on manpower, the production period is often long, and the hysteresis of the electronic map is serious. With the increasing demand of the public on the electronic map, the short board of the traditional operation method in the aspect of the situation is more serious. Therefore, in order to solve this problem, it is necessary to provide a method for customizing map style drawings.

Disclosure of Invention

The embodiment of the invention provides a drawing method, a drawing system, a storage medium and drawing equipment for a user-defined map style, which can support the setting of the user-defined map style.

According to the embodiment of the invention, the drawing method of the custom map style is provided, and comprises the following steps:

training a convolutional neural network classification model based on the image sample set of the picture;

extracting a target ground object in the image based on the model of the convolutional neural network, and calculating the target ground object to obtain a grid data set for ground object classification in the picture;

vectorizing the grid data set to obtain vector image elements after vectorization;

and editing the vector image elements by inputting map parameters to complete the map making of the custom style so as to output a map of the custom style.

Further, before editing the vector image elements by inputting map parameters to complete the map drawing of the custom style so as to output a map of the custom style, the method further comprises the following steps:

and selecting a preset map style to output a preset style map.

Further, based on the image sample of the picture, the model for training the convolutional neural network is specifically as follows:

collecting picture image samples to generate a picture image sample set;

analyzing the picture image sample set into training data according to a uniform format;

and combining and training the training data to generate a model for training the convolutional neural network.

Further, extracting a target ground object in the image based on the model of the convolutional neural network, and calculating the target ground object to obtain a grid data set for ground object classification in the picture, wherein the grid data set specifically comprises:

applying the trained convolutional neural network classification model to a new data area for ground object classification, and extracting a target ground object in the picture;

and obtaining a grid data set of the ground object classification of the picture based on the target ground object.

Further, vectorizing the raster data set to obtain vector image elements after vectorization specifically includes:

graying the color image of the grid data set to obtain a gray image, and performing binarization processing on the gray image;

carrying out noise reduction processing on the gray level image subjected to binarization processing;

thinning the image subjected to noise reduction processing, and extracting a main skeleton of the image;

converting the extracted main skeleton into a coordinate sequence of a vector graph, and carrying out topological processing on the coordinate sequence of the vector graph to obtain vectorized data;

and searching unreasonable data for the vectorized data, and deleting the unreasonable data to obtain vector image elements with the unreasonable data deleted.

Further, finding out unreasonable data for the vectorized data, and deleting the unreasonable data to obtain vector image elements from which the unreasonable data are deleted, the method further includes:

and comparing the vector image elements after vectorization with the same positions of the grid images, and judging the accuracy and the reasonable degree of the vectorized image.

A custom map style mapping system, comprising:

the model training module is used for training a model of the convolutional neural network based on the picture image sample set;

the target extraction module is used for extracting a target ground object in the image based on the model of the convolutional neural network, calculating the target ground object and obtaining a grid data set for ground object classification in the image;

the rasterization module is used for carrying out vectorization processing on the raster data set to obtain vector image elements after vectorization processing;

and the style setting module is used for editing the vector image elements by inputting map parameters to complete the map making of the custom style so as to output a custom style map.

Further, the system further comprises:

and the style selection module is used for selecting the style of the preset map so as to output the preset style map.

A computer readable storage medium, wherein the computer readable storage medium stores one or more programs, the one or more programs being executable by one or more processors to perform steps in a customized map-style mapping method as in any above.

A terminal device, comprising: a processor, a memory, and a communication bus; the memory has stored thereon a computer readable program executable by the processor;

the communication bus realizes the connection communication between the processor and the memory;

the processor executes the computer readable program to realize the steps of the customized map style drawing method.

In the method, the system, the storage medium and the equipment for drawing the customized map style, the method comprises the following steps: training a model of a convolutional neural network based on the image sample set of the picture; applying the trained convolutional neural network classification model to a new data area to classify the ground objects to obtain a grid data set of ground object classification in the picture; vectorizing the obtained raster data set, and obtaining a vector diagram of each element; and editing the vector diagram by inputting map parameters to complete the map making of the custom style so as to output the custom style map. The method and the device solve the problem that the user is difficult to obtain the map with the custom style, solve a plurality of limitations of the map with the custom style, and enable the user to customize the map with the style according to the preference of the user.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:

FIG. 1 is a flow chart of a custom map style mapping method of the present invention;

FIG. 2 is a schematic diagram of a custom map style mapping system of the present invention;

FIG. 3 is an AI inference diagram of the present invention;

FIG. 4 is a schematic diagram of object and background binarization for pixels in accordance with the present invention;

FIG. 5 is a schematic diagram of image preprocessing according to the present invention;

FIG. 6 is a schematic view of the present invention showing the reduction of noise and stain removal;

FIG. 7 is a schematic diagram of a subject skeleton from which an image is extracted according to the present invention;

FIG. 8 is a schematic diagram of a coordinate sequence for converting a skeleton into a vector graphic according to the present invention;

fig. 9 is a schematic diagram of a terminal device provided in the present invention.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention. The embodiments described by referring to the drawings are exemplary only for the purpose of illustrating the invention and are not to be construed as limiting the invention. In addition, if a detailed description of the known art is not necessary to show the features of the present invention, it is omitted.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

Referring to fig. 1, the present application provides a customized map style drawing method, including the following steps:

s101: training a convolutional neural network classification model based on the image sample set of the picture;

s102: extracting a target ground object in the image based on the model of the convolutional neural network, and calculating the target ground object to obtain a grid data set for ground object classification in the picture;

s103: vectorizing the grid data set to obtain vector image elements after vectorization;

s104: and editing the vector image elements by inputting map parameters to complete the map making of the custom style so as to output a map of the custom style.

In an embodiment, the customized map style drawing method of the invention includes: training a model of a convolutional neural network based on the image sample set of the picture; applying the trained convolutional neural network classification model to a new data area to classify the ground objects to obtain a grid data set of ground object classification in the picture; vectorizing the obtained raster data set, and obtaining a vector diagram of each element; and editing the vector diagram by inputting map parameters to complete the map making of the custom style so as to output the custom style map. The method and the device solve the problem that the user is difficult to obtain the map with the custom style, solve a plurality of limitations of the map with the custom style, and enable the user to customize the map with the style according to the preference of the user.

In an embodiment, before the editing the vector image element by inputting the map parameter to complete the map making in the custom style so as to output the map in the custom style, the method further includes:

and selecting a preset map style to output a preset style map.

The method comprises the steps that templates with various map styles are preset in the application and can be selected by a user, the preset templates are preset map parameters which are input in advance, a designed map wind template is generated according to the preset parameters, the user can select favorite map styles from the preset templates and directly use the favorite map styles, and therefore the step of user design is omitted.

In the embodiment, based on the image sample, the model for training the convolutional neural network specifically comprises:

collecting picture image samples to generate a picture image sample set;

The following is the specific steps of the classification model training process of the convolutional neural network,

the method comprises the following steps: a large number of image samples are collected.

Specifically, the data acquisition approach can be from a national geographic information spatio-temporal cloud platform or from an open geographic service api.

Step two: and (4) preprocessing data. Namely, the acquired image samples are stored and analyzed according to a uniform format, and the analyzed image samples are combined into training data.

Specifically, a large number of samples are needed for model training, the more samples are, the more accurate the trained model is, and when the number of samples is insufficient, the samples can be enhanced by adopting rotation or symmetrical change so as to improve the efficiency of model training.

Step three: and (5) training data is arranged, and the training data is used as neural network input.

Specifically, after the training data set is prepared, a neural network propagation process and model parameters are defined, a model is trained on the basis of training data, the model is evaluated on verification data, optimal parameters are searched in the training model and model evaluation processes, once the optimal parameters are found, the test data is tested for the last time, and the model training process is completed.

Step four: and testing the verification model.

Specifically, after the model training process is completed, the model can be tested to enable the model to reach the optimal precision, the training process is only needed to be carried out once when the model is built, the optimal model can be found, and the model can be reused.

In the embodiment, the extracting of the target ground object in the image based on the model of the convolutional neural network, and the calculating of the target ground object to obtain the grid data set of the ground object classification in the picture are specifically as follows:

Specifically, AI reasoning is performed on the model after training of the model is completed. The AI reasoning refers to a process of applying the trained convolutional neural network classification model to a new data area for ground object classification, so as to rapidly extract a target ground object, and finally obtain a ground object classification labeling result of the picture.

The number of layers of a neural network and the number of neurons in each layer are arbitrary according to the neural network algorithm. The basic logic is the same: the input is transmitted forward in the neural network, and finally the output is obtained. As shown in fig. 3, the middle are all the influence factors, or weights; in the process of sample identification, the weight is continuously adjusted, so that the final calculation result is fitted with a correct result as much as possible. When a certain accuracy is reached, the AI inference algorithm is finished. And after the AI inference algorithm is completed, calculating the target sample to finally obtain a ground object classification grid data result.

In an embodiment, vectorizing the grid data set to obtain vector image elements after vectorization specifically includes:

Specifically, vectorizing raster data obtained after reasoning the model AI; the following is a detailed description of the grid data vectorization:

the first step is as follows: and carrying out binarization processing on the image.

Specifically, referring to fig. 4, a grayed image is obtained by graying a color image (image graying refers to an image having only one sample color per pixel, such an image is generally displayed as a gray scale from darkest black to brightest white); and carrying out binarization processing on the acquired gray level image. The most common method for binarization of grayscale images is a threshold method, i.e. the difference between the target and the background in the image is utilized to set the image into two different levels, and a suitable threshold is selected, such as: 0-126, 255 to determine whether a pixel is a target or a background, thereby obtaining a binarized image.

The second step is that: and carrying out noise reduction pretreatment on the gray level image after the binarization treatment.

Specifically, referring to fig. 5, some blooming, stain, streaking edge irregularity, etc. always occur due to image noise or other reasons. The importance of image noise in digital image processing technology is more and more obvious, and interpretation, noise removal and the like of high-magnification aerial photographs become indispensable technical steps. At this time, the raster data is preprocessed by the following method:

for example, in the case where both the flying white and the black dots occur, the peripheral adjacent gradation values are uniform, and in the case where an abnormal value occurs, the flying white and the black dots need to be corrected. Image noise may be denoised using an image processing tool such as ENVI. The noise reduction processing mode generally selects a median filter for noise reduction.

The median filter is a common nonlinear smoothing filter, and the basic principle is to substitute the value of one point in a digital image or a digital sequence by the median of each point value in the field of the point, and the median filter has the main function of changing the pixel with larger difference of the gray values of the surrounding pixels into the value close to the value of the surrounding pixels, so that an isolated noise point can be eliminated, and the median filter is very effective for filtering the salt and pepper noise of the image (the salt and pepper noise is the pixel with black and white randomly appearing on the image). The median filter can remove noise and protect the edge of the image, thereby obtaining more satisfactory restoration effect.

The third step: and thinning the image after the noise reduction is finished, and extracting a main skeleton or skeleton of the image.

Specifically, referring to FIG. 6, a and b are peelable, while the c and d center points are non-peelable; and gradually stripping points on the edge of the outline of the preprocessed pixel array to form a skeleton pattern with the width of a line being only one pixel. The thinned graph skeleton not only retains most of the characteristics of the original graph, but also is convenient for the next processing.

Wherein, the basic process of the refinement is as follows:

(1) determining a pixel set needing to be refined;

(2) removing picture elements that are not skeletons;

(3) repeating the above steps until only skeleton pixel is left.

Further, referring to fig. 7, the refined data should satisfy the following requirements: (1) the continuity of the original line is kept; (2) the line width is only one pixel; (3) the skeleton after fine drawing is the central line of the original line drawing; (4) the original features of the pattern are preserved.

The fourth step: and tracking the coordinate sequence of the vector graphics to obtain vectorized data.

Specifically, referring to fig. 8, the tracking is to convert the skeleton into a coordinate sequence of a vector graph, and finally find the end points, nodes, and isolated points of a line, perform the topology to obtain vectorized data, obtain a vector graph of each element, and perform the vector data processing. And vector data processing is carried out after the vector diagram of each element is obtained. Wherein, the basic steps of tracking are as follows:

(1) searching lines from left to right and from top to bottom to draw a starting point and recording coordinates;

(2) tracking the point towards 8 directions of the point, if not, ending the tracking of the current line, and turning to (1) step-by-step tracking of the next line; otherwise, recording the coordinates;

(3) and (4) moving the search point to the newly-taken point, and repeating the step (2).

The fifth step: the vectorized data is subjected to data processing,

specifically, unreasonable elements such as extremely small face elements, topological error elements, multi-face elements and the like can be processed, and the unreasonable elements can be found out through a spatial analysis algorithm, and particularly can be found out through arcgis or other geographic software tools according to topological relations and attributes among data; for example, an element having an area smaller than 0.01 square meter may be calculated, or an element having a surface overlapping with a surface may be found from the positional relationship.

Finding out unreasonable data and then processing the unreasonable data; the vector elements meeting the updating requirement of the electronic map are obtained by deleting extremely small face elements (the area of a face element is smaller than a certain threshold value and is generally defined to be smaller than 0.01 square meter), processing vector topological errors (such as face overlapping, gaps between faces and the like), eliminating multi-face data (one face element comprises a plurality of faces), performing polygon boundary smoothing processing and the like by using a software tool.

Further, the correction of the boundary and topological relation mainly corrects errors such as overlapping of suspension lines and elements and unclosed surfaces, for example, correction and improvement are required when administrative regions are overlapped and main rivers are cut off, and the improvement of the attributes is to supplement, delete and improve the missing attributes or the error attributes of the elements.

In the embodiment, after finding out the unreasonable data for the vectorized data and deleting the unreasonable data to obtain the vector image elements from which the unreasonable data is deleted, the method further includes:

Specifically, after the fifth step, overlay analysis may be performed, where the overlay analysis mainly compares the vectorized image with the grid image at the same position, and human eyes may be determined according to the vectorized accuracy and the vectorized reasonable degree.

Further, after vectorization processing is performed on the raster data set to obtain vector image elements after vectorization processing, the vectorized elements are obtained, and a user can input assignment to output custom map tile data.

Specifically, the user can edit vector elements, edit colors and patterns of the elements by using an editor, assign values to styles input by the user, such as various colors and layer matching of water systems, land and buildings, match colors by using a relatively complete symbol library in the software, and customize the style by the user.

After the style is set, the scale and level are set (typically using google map scale and level parameters). And then, carrying out plotting processing to finish the graphic output of the target ground object, thereby finishing the map drawing with the custom style. In addition, the template has various forms, and can be selected by a user, so that the step of designing by the user is omitted.

The customized map style drawing method based on deep learning meets the requirements of a user on the basis of meeting the technical requirements of an electronic map, and compared with the traditional modes of field incremental information acquisition, interior work change information extraction and interior and exterior work coordination obtaining change information, the change information extraction based on deep learning can greatly shorten the processing period and reduce the manual workload by more than 80%.

Compared with the prior art, the method and the device can support use in any network environment, and support any style, and the style is set by a user in a self-defining way. The method and the device solve a plurality of limitations and solve the problem that the user is difficult to obtain the map with the custom style. The geometric accuracy of the geographic elements extracted through deep learning can meet the technical requirements of an electronic map, the change area of the earth surface is found quickly through spatial analysis, and a large amount of element acquisition workload can be saved. In the area without ready-made data, the updating efficiency of the self-defined electronic map is greatly improved, and the method has great significance for the quick updating of the electronic map.

Referring to fig. 2, the present application provides a mapping system with customized map style, comprising:

the model training module 100 is used for training a model of a convolutional neural network based on the picture image sample set;

the target extraction module 200 is configured to extract a target feature in the image based on a convolutional neural network model, and calculate the target feature to obtain a grid data set for classifying the feature in the image;

the rasterizing module 300 is configured to perform vectorization processing on the raster data set to obtain vector image elements after vectorization processing;

the style setting module 400 edits the vector image elements by inputting map parameters to complete the map making of the custom style so as to output a map of the custom style.

In an embodiment, the system further comprises:

In the customized map style drawing method, system, storage medium and device of the present invention, the system comprises: the model training module is used for training a model of the convolutional neural network based on the picture image sample set; the target extraction module is used for carrying out ground feature classification by applying the trained convolutional neural network classification model to a new data area to obtain a grid data set of ground feature classification in the picture; the rasterization module is used for carrying out vectorization processing on the obtained raster data set and obtaining a vector diagram of each element; and the style setting module is used for editing the vector diagram by inputting map parameters to complete the map making of the custom style so as to output the custom style map. The method and the device solve the problem that the user is difficult to obtain the map with the custom style, solve a plurality of limitations of the map with the custom style, and enable the user to customize the map with the style according to the preference of the user.

Based on the customized map style mapping method, the embodiment provides a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the customized map style mapping method according to the embodiment.

Based on the mapping method of the customized map style, the present application further provides a terminal device, as shown in fig. 9, which includes at least one processor (processor) 20; a display screen 21; and a memory (memory)22, and may further include a communication interface (communication interface)23 and a bus 24. The processor 20, the display 21, the memory 22 and the communication interface 23 can communicate with each other through the bus 24. The display screen 21 is configured to display a user guidance interface preset in the initial setting mode. The communication interface 23 may transmit information. Processor 20 may invoke logic instructions in memory 22 to perform the custom map-style mapping method of the above-described embodiment.

Furthermore, the logic instructions in the memory 22 may be implemented in software functional units and stored in a computer readable storage medium when sold or used as a stand-alone product.

The memory 22, which is a computer-readable storage medium, may be configured to store a software program, a computer-executable program, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure. The processor 20 executes the functional application and data processing, i.e. implements the method in the above-described embodiments, by executing the software program, instructions or modules stored in the memory 22.

The memory 22 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal device, and the like.

Further, the memory 22 may include a high speed random access memory and may also include a non-volatile memory. For example, various media that can store program codes, such as a usb disk, a removable hard disk, a Read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, may also be transient storage media.

In addition, the specific processes loaded and executed by the storage medium and the instruction processors in the terminal device are described in detail in the method, and are not stated herein.

While the foregoing is directed to the preferred embodiment of the present invention, and the specific embodiments of the present invention are not limited to the foregoing description, it will be appreciated by those skilled in the art that various changes may be made without departing from the principles of the invention, and that such changes and modifications are to be considered as within the scope of the invention.

Claims

1. The drawing method of the custom map style is characterized by comprising the following steps of:

2. The customized mapping method according to claim 1, wherein before the editing the vector image elements by inputting map parameters to complete customized mapping, the method further comprises:

and selecting a preset map style to output a preset style map.

3. The custom map style drawing method according to claim 1, wherein the model for training the convolutional neural network based on the picture image sample is specifically:

collecting picture image samples and generating a picture image sample set;

and combining and training the training data to generate a model for training a convolutional neural network.

4. The custom map style drawing method according to claim 1, wherein the extracting a target feature in an image based on the convolutional neural network model, and calculating the target feature to obtain a grid data set for classifying the feature in the image specifically comprises:

5. The custom map style drawing method according to claim 1, wherein the vectorizing the grid data set to obtain vectorized vector image elements specifically comprises:

graying the color image of the raster data set to obtain a gray level image, and performing binarization processing on the gray level image;

carrying out noise reduction processing on the gray level image after binarization processing;

6. The customized mapping style mapping method according to claim 5, wherein after searching for unreasonable data for the vectorized data and deleting the unreasonable data to obtain vector image elements with unreasonable data deleted, the method further comprises:

7. A custom map style mapping system, comprising:

the target extraction module is used for extracting a target ground object in the image based on the model of the convolutional neural network, and calculating the target ground object to obtain a grid data set for ground object classification in the image;

and the style setting module is used for editing the vector image elements by inputting map parameters to complete map making of a user-defined style so as to output a map of the user-defined style.

8. The custom map style mapping system of claim 7, further comprising:

9. A computer readable storage medium, storing one or more programs, the one or more programs being executable by one or more processors to perform the steps in the custom map style mapping method of any of claims 1-6.

10. A terminal device, comprising: a processor, a memory, and a communication bus; the memory has stored thereon a computer readable program executable by the processor;

the communication bus realizes connection communication between the processor and the memory;

the processor, when executing the computer readable program, performs the steps of the method of customized mapping of a map style of any of claims 1-6.