An intelligent drawing system and method based on AI

Abstract

The present invention belongs to the field of AI drawing and relates to data analysis technology, which is used to solve the problem that the intelligent drawing system in the prior art occupies drawing resources but cannot improve user experience. Specifically, it is an AI-based intelligent drawing system and method, including an intelligent drawing platform, which is communicatively connected with an original image processing module, a feature analysis module, a drawing execution module and a database; the original image processing module is used to perform feature processing and analysis on the original image of the intelligent drawing and obtain feature parameters of the original image, and send the feature parameters of the original image to the feature analysis module through the intelligent drawing platform; the present invention can perform feature processing and analysis on the original image of the intelligent drawing, compare and analyze the historical feature parameters of the historical drawing data with the feature parameters of the original image, and obtain a calibrated original image and a calibrated generated image, so that the correct painting decision can be used to perform AI painting processing on the original image.

Description

Technical Field

[0001] The present invention belongs to the field of AI drawing, relates to data analysis technology, and specifically is an AI-based intelligent drawing system and method. Background Art

[0002] Artificial intelligence painting breaks through the limits of human beings themselves, allowing painting analysis to enter a broader perspective with humanistic spirit as the starting point and end point; through artificial intelligence, a new field of painting art is opened up.

[0003] The intelligent drawing system in the prior art can only extract features from the original image and then use fixed processing methods to generate Ai images. However, when the application scenarios of Ai drawing are different, the processing methods required are also different. The processing solutions in the prior art generally use multiple processing methods to obtain multiple Ai drawing results and then screen them, or directly allow users to choose from multiple Ai drawing results. However, this method has a low utilization rate of Ai drawing tools, and the drawing results generated for users are less targeted. Therefore, there is a problem of occupying drawing resources but failing to improve user experience.

[0004] In view of the above technical problems, this application proposes a solution. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent drawing system and method based on Ai, which is used to solve the problem that the intelligent drawing system in the prior art occupies drawing resources but cannot improve user experience;

[0006] The technical problem to be solved by the present invention is: how to provide an AI-based intelligent drawing system and method that can improve the utilization rate of AI drawing tools and improve user experience.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] An AI-based intelligent drawing system includes an intelligent drawing platform, wherein the intelligent drawing platform is communicatively connected with an original image processing module, a feature analysis module, a drawing execution module and a database;

[0009] The original image processing module is used to perform feature processing and analysis on the original image of the intelligent drawing and obtain feature parameters of the original image, and send the feature parameters of the original image to the feature analysis module through the intelligent drawing platform;

[0010] The feature analysis module is used to perform feature processing and analysis on the original image of the intelligent drawing and mark the drawing strategy of the original image as basic drawing, added drawing or deleted drawing; the drawing strategy of the original image is sent to the drawing execution module through the intelligent drawing platform;

[0011] The drawing execution module is used to execute the intelligent drawing task according to the drawing strategy of the original image: perform range calculation on the feature parameters of the original image to obtain the feature parameter range, and mark the historical feature parameters e and the historical drawing data e corresponding to the feature parameter range as reference data; when the drawing strategy of the original image is marked as basic drawing, execute step K1; when the drawing strategy of the original image is marked as subtractive drawing, execute steps K2-K1; when the drawing strategy of the original image is marked as additive drawing, execute steps K3-K1; when the drawing strategy of the original image is marked as both subtractive drawing and additive drawing, execute steps K2-K3-K1.

[0012] Furthermore, the process of acquiring the feature parameters of the original image includes: generating drawing path information according to the original image, the drawing path information includes several groups, each group includes several anchor points, marking the group with the number of anchor points not less than L1 as a feature group, marking the number of feature groups as the rich data FFy of the original image, connecting the anchor points in the feature group to obtain the line data of the feature group, calculating the smoothness of the line data, summing and averaging the smoothness of the line data of all feature groups to obtain the smooth data PHy of the original image, performing grayscale transformation on the original image to obtain a grayscale image, marking the average value of the grayscale values ​​of the corresponding areas of all feature groups in the grayscale image as the color data YSy of the original image; the rich data FFy, the smoothed data PHy and the color data YSy constitute the feature parameters of the original image.

[0013] Furthermore, the feature analysis module is used to perform feature processing and analysis on the original image of the intelligent drawing. The specific process includes: retrieving the historical drawing data e of the last L2 months through the database, where e=1, 2, ..., n, where n is a positive integer, and the historical drawing data e includes the historical original image e, the historical feature parameter e, the historical Ai painting work e and the Ai processing means e; calculating the conformity coefficient FHe of the historical drawing data e relative to the original image; marking the historical original image e and the historical Ai painting work e of the historical drawing data e with the smallest conformity coefficient FHe as the calibrated original image and the calibrated generated image, respectively, extracting the elements of the calibrated original image and the calibrated generated image and establishing the calibrated original element set and the calibrated generated element set, respectively; comparing the calibrated original element set with the calibrated generated element set and marking the drawing strategy of the original image through the comparison result.

[0014] Furthermore, the process of obtaining the conformity coefficient FHe of the historical drawing data e relative to the original image includes: comparing and analyzing the rich data FFe, smoothing data PHe and color data YSe of the historical feature parameters e with the feature parameters of the original image: by formula The conformity coefficient FHe of the historical drawing data e is obtained, where m1, m2 and m3 are all proportional coefficients, and m1>m2>m3>1.

[0015] Furthermore, the specific process of comparing the calibrated original element set with the calibrated generated element set includes: if the element types in the calibrated original element set and the calibrated generated element set are exactly the same, the drawing strategy of the original image is marked as basic drawing; if the calibrated original element set contains elements outside the calibrated generated element set, the drawing strategy of the original image is marked as deleted drawing; if the calibrated generated element set contains elements outside the calibrated original element set, the drawing strategy of the original image is marked as added drawing.

[0016] Furthermore, the process of obtaining the characteristic parameter range includes: obtaining the original rich low value FFmin and the original rich high value FFmax through the formula FFmin=p1*FFy and FFmax=p2*FFy, where p1 and p2 are both proportional coefficients, and 0.85≤p1≤0.95; 1.05≤p2≤1.15; the rich range is composed of the original rich low value FFmin and the original rich high value FFmax; obtaining the original smooth low value PHmin and the original smooth high value PHmax through the formula PHmin=p3*PHy and PHmax=p4*PHy, where p3 and p4 are both proportional coefficients Coefficient, and 0.85≤p3≤0.95; 1.05≤p4≤1.15; the smoothing range is composed of the original smoothing low value PHmin and the original smoothing high value PHmax; the original color low value YSmin and the original color high value YSmax are obtained by the formulas YSmin=p5*YSy and YSmax=p6*YSy, where p5 and p6 are both proportional coefficients, and 0.85≤p5≤0.95; 1.05≤p6≤1.15; the color range is composed of the original color low value YSmin and the original color high value YSmax; the feature parameter range is composed of the rich range, smoothing range and color range.

[0017] Furthermore, the specific process of performing the intelligent painting task according to the drawing strategy of the original image includes the following steps:

[0018] Step K1: retrieve the Ai processing means e corresponding to the historical drawing data e with the smallest coefficient FHe value and mark it as the priority means, and use the Ai painting tool to process the original image using the priority means to obtain the Ai painting work;

[0019] Step K2: Compare the original image in the reference data with the generated image: mark the image elements that are exclusive to the original image as deleted elements, count the deleted elements of all reference data, mark the number of marks corresponding to the deleted elements as the elimination value of the deleted elements, mark the L2 deleted elements with the largest elimination value as the deletion execution elements, and eliminate the deletion execution elements in the drawing task;

[0020] Step K3: Compare the original image in the reference data with the generated image: mark the image elements that are unique to the generated image as added elements, count the added elements of all reference data, mark the number of markings corresponding to the added elements as the added values ​​of the added elements, mark the L2 added elements with the largest added values ​​as added execution elements, and add the added execution elements to the drawing task.

[0021] An AI-based intelligent drawing method comprises the following steps:

[0022] Step 1: Perform feature processing and analysis on the original image of the intelligent drawing and obtain the feature parameters of the original image;

[0023] Step 2: Perform feature processing and analysis on the original image of the intelligent drawing and mark the drawing strategy of the original image;

[0024] Step 3: Execute the intelligent painting task according to the drawing strategy of the original image to obtain the Ai painting work.

[0025] The present invention has the following beneficial effects:

[0026] 1. The original image processing module can be used to perform feature processing and analysis on the original image of intelligent drawing, and the feature parameters of the original image can be mined and extracted from different angles, and the feature parameters can be used to provide feedback on the drawing features of the original image;

[0027] 2. The feature analysis module can be used to perform feature processing and analysis on the original image of intelligent drawing, compare and analyze the historical feature parameters of the historical drawing data with the feature parameters of the original image, and obtain the calibrated original image and the calibrated generated image, so as to mark the painting strategy of the original image according to the calibrated original image and the calibrated generated image, and then the correct painting decision can be used to perform Ai painting on the original image;

[0028] 3. The drawing execution module is used to execute intelligent painting tasks according to the drawing strategy of the original image. After the corresponding increase and decrease processing of the painting elements in the original image, the original image is processed according to the corresponding priority means to improve the utilization rate of Ai painting tools and the pertinence of the generated results. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 is a system block diagram of Embodiment 1 of the present invention;

[0031] Figure 2 This is a flow chart of the method of Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0032] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Embodiment 1: Figure 1 As shown, an AI-based intelligent drawing system includes an intelligent drawing platform, which is communicatively connected to an original image processing module, a feature analysis module, a drawing execution module and a database.

[0034] The original image processing module is used to perform feature processing and analysis on the original image of intelligent drawing: drawing path information is generated according to the original image, the drawing path information includes several groups, each group includes several anchor points, and the group with the number of anchor points not less than L1 is marked as a feature group, and the number of feature groups is marked as the rich data FFy of the original image, the anchor points in the feature group are connected to obtain the line data of the feature group, the smoothness of the line data is calculated, the smoothness of the line data of all feature groups is summed and averaged to obtain the smooth data PHy of the original image, the original image is gray-scale transformed to obtain a gray-scale image, and the average value of the gray-scale values ​​of all feature groups in the corresponding areas of the gray-scale image is marked as the color data YSy of the original image; the rich data FFy, the smooth data PHy and the color data YSy constitute the feature parameters of the original image, and the feature parameters of the original image are sent to the feature analysis module through the intelligent drawing platform.

[0035] The feature analysis module is used to perform feature processing and analysis on the original image of intelligent drawing: retrieve the historical drawing data e of the last L2 months through the database, e = 1, 2, ..., n, n is a positive integer, and the historical drawing data e includes historical original images e, historical feature parameters e, historical Ai paintings e and Ai processing methods e; compare and analyze the rich data FFe, smoothing data PHe and color data YSe in the historical feature parameters e with the feature parameters of the original image: by formula Obtain the conformity coefficient FHe of the historical drawing data e, wherein m1, m2 and m3 are all proportional coefficients, and m1>m2>m3>1; mark the historical original image e and the historical Ai painting e of the historical drawing data e with the smallest conformity coefficient FHe as the calibrated original image and the calibrated generated image respectively, extract the elements of the calibrated original image and the calibrated generated image, and establish the calibrated original element set and the calibrated generated element set respectively; compare the calibrated original element set with the calibrated generated element set: if the element types in the calibrated original element set and the calibrated generated element set are exactly the same, mark the drawing strategy of the original image as basic drawing; if the calibrated original element set contains elements outside the calibrated generated element set, mark the drawing strategy of the original image as deleted drawing; if the calibrated generated element set contains elements outside the calibrated original element set, mark the drawing strategy of the original image as added drawing; send the drawing strategy of the original image to the drawing execution module through the intelligent drawing platform.

[0036] The drawing execution module is used to execute the intelligent drawing task according to the drawing strategy of the original image: the feature parameter range is obtained by range calculation of the feature parameters of the original image: the original rich low value FFmin and the original rich high value FFmax are obtained by the formula FFmin=p1*FFy and FFmax=p2*FFy, wherein p1 and p2 are both proportional coefficients, and 0.85≤p1≤0.95; 1.05≤p2≤1.15; the rich range is formed by the original rich low value FFmin and the original rich high value FFmax; the original smooth low value PHmin and the original smooth high value PHmax are obtained by the formula PHmin=p3*PHy and PHmax=p4*PHy, wherein p3 and p4 are both proportional coefficients, and 0.85≤p3≤0.95; 1.05≤p4≤1.15; the smooth range is formed by the original smooth low value PHmin and the original smooth high value PHmax. ; The original color low value YSmin and the original color high value YSmax are obtained by the formula YSmin=p5*YSy and YSmax=p6*YSy, wherein p5 and p6 are both proportional coefficients, and 0.85≤p5≤0.95; 1.05≤p6≤1.15; The original color low value YSmin and the original color high value YSmax constitute a color range; The rich range, the smooth range and the color range constitute a feature parameter range; The historical feature parameter e and the historical drawing data e corresponding to the feature parameter range are marked as reference data; When the drawing strategy of the original image is marked as basic drawing, step K1 is executed; When the drawing strategy of the original image is marked as subtractive drawing, steps K2-K1 are executed; When the drawing strategy of the original image is marked as additive drawing, steps K3-K1 are executed; When the drawing strategy of the original image is marked as both subtractive drawing and additive drawing, steps K2-K3-K1 are executed;

[0037] The specific process of performing the intelligent painting task according to the drawing strategy of the original image includes the following steps:

[0038] Step K1: retrieve the Ai processing means e corresponding to the historical drawing data e with the smallest coefficient FHe value and mark it as the priority means, and use the Ai painting tool to process the original image using the priority means to obtain the Ai painting work;

[0039] Step K2: Compare the original image in the reference data with the generated image: mark the image elements that are exclusive to the original image as deleted elements, count the deleted elements of all reference data, mark the number of marks corresponding to the deleted elements as the elimination value of the deleted elements, mark the L2 deleted elements with the largest elimination value as the deletion execution elements, and eliminate the deletion execution elements in the drawing task;

[0040] Step K3: Compare the original image in the reference data with the generated image: mark the image elements that are unique to the generated image as added elements, count the added elements of all reference data, mark the number of markings corresponding to the added elements as the added values ​​of the added elements, mark the L2 added elements with the largest added values ​​as added execution elements, and add the added execution elements to the drawing task.

[0041] Embodiment 2: Figure 2 As shown, an intelligent drawing method based on AI comprises the following steps:

[0042] Step 1: Perform feature processing and analysis on the original image of the intelligent drawing and obtain the feature parameters of the original image;

[0043] Step 2: Perform feature processing and analysis on the original image of the intelligent drawing and mark the drawing strategy of the original image;

[0044] Step 3: Execute the intelligent painting task according to the drawing strategy of the original image to obtain the Ai painting work.

[0045] An AI-based intelligent drawing system and method, when working, performs feature processing and analysis on the original image of the intelligent drawing and obtains the feature parameters of the original image; sends the feature parameters of the original image to the feature analysis module through the intelligent drawing platform; performs feature processing and analysis on the original image of the intelligent drawing and marks the drawing strategy of the original image; sends the drawing strategy of the original image to the drawing execution module through the intelligent drawing platform; executes the intelligent drawing task according to the drawing strategy of the original image to obtain the Ai painting work.

[0046] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

[0047] The above formulas are obtained by collecting a large amount of data and performing software simulation to select a formula close to the actual value. The coefficients in the formula are set by technicians in this field according to actual conditions; for example: Formula A technician in this field collects multiple groups of sample data and sets a corresponding compliance coefficient for each group of sample data; substitutes the set compliance coefficient and the collected sample data into the formula, and any three formulas constitute a three-variable linear equation system. The calculated coefficients are screened and averaged, and the values ​​of m1, m2, and m3 are obtained to be 2.86, 2.31, and 1.62, respectively;

[0048] The size of the coefficient is to quantify each parameter to obtain a specific value for subsequent comparison. The size of the coefficient depends on the amount of sample data and the technical personnel in this field preliminarily setting the corresponding compliance coefficient for each set of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0049] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0050] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An AI-based intelligent mapping system, characterized in that: It includes an intelligent drawing platform, which is communicatively connected with an original image processing module, a feature analysis module, a drawing execution module and a database; The original image processing module is used to perform feature processing and analysis on the original image of the intelligent drawing and obtain feature parameters of the original image, and send the feature parameters of the original image to the feature analysis module through the intelligent drawing platform; The feature analysis module is used to perform feature processing and analysis on the original image of the intelligent drawing and mark the drawing strategy of the original image as basic drawing, added drawing or deleted drawing; the drawing strategy of the original image is sent to the drawing execution module through the intelligent drawing platform; The drawing execution module is used to execute the intelligent drawing task according to the drawing strategy of the original image: perform range calculation on the feature parameters of the original image to obtain the feature parameter range, and mark the historical feature parameter e and the historical drawing data e corresponding to the feature parameter range as reference data; when the drawing strategy of the original image is marked as basic drawing, execute step K1; when the drawing strategy of the original image is marked as deletion drawing, execute steps K2-K1; when the drawing strategy of the original image is marked as addition drawing, execute steps K3-K1; when the drawing strategy of the original image is marked as both deletion drawing and addition drawing, execute steps K2-K3-K1; The process of acquiring the characteristic parameters of the original image includes: generating drawing path information according to the original image, the drawing path information includes a number of groups, each group includes a number of anchor points, marking the group with the number of anchor points not less than L1 as a characteristic group, marking the number of characteristic groups as the rich data FFy of the original image, connecting the anchor points in the characteristic group to obtain the line data of the characteristic group, calculating the smoothness of the line data, summing and averaging the smoothness of the line data of all the characteristic groups to obtain the smooth data PHy of the original image, performing grayscale transformation on the original image to obtain a grayscale image, marking the average value of the grayscale values ​​of the corresponding areas of all the characteristic groups in the grayscale image as the color data YSy of the original image; the characteristic parameters of the original image are constituted by the rich data FFy, the smooth data PHy and the color data YSy; The specific process of the feature analysis module for feature processing and analysis of the original image of intelligent drawing includes: retrieving the historical drawing data e of the last L2 months through the database, e=1, 2, ..., n, n is a positive integer, and the historical drawing data e includes the historical original image e, the historical feature parameter e, the historical Ai painting work e and the Ai processing means e; calculating the conformity coefficient FHe of the historical drawing data e relative to the original image; marking the historical original image e and the historical Ai painting work e of the historical drawing data e with the smallest conformity coefficient FHe as the calibration original image and the calibration generated image respectively, extracting the elements of the calibration original image and the calibration generated image and establishing the calibration original element set and the calibration generated element set respectively; comparing the calibration original element set with the calibration generated element set and marking the drawing strategy of the original image through the comparison result; The process of obtaining the conformity coefficient FHe of the historical drawing data e relative to the original image includes: comparing and analyzing the rich data FFe, smoothing data PHe and color data YSe of the historical feature parameters e with the feature parameters of the original image: Obtain the coincidence coefficient FHe of the historical drawing data e, where m1, m2 and m3 are all proportional coefficients, and m1>m2>m3>1; The specific process of comparing the calibration original element set with the calibration generated element set includes: if the element types in the calibration original element set and the calibration generated element set are exactly the same, then the drawing strategy of the original image is marked as basic drawing; if the calibration original element set contains elements outside the calibration generated element set, then the drawing strategy of the original image is marked as deleted drawing; if the calibration generated element set contains elements outside the calibration original element set, then the drawing strategy of the original image is marked as added drawing; The specific process of performing the intelligent painting task according to the drawing strategy of the original image includes the following steps: Step K1: retrieve the Ai processing means e corresponding to the historical drawing data e with the smallest coefficient FHe value and mark it as the priority means, and use the Ai painting tool to process the original image using the priority means to obtain the Ai painting work; Step K2: Compare the original image in the reference data with the generated image: mark the image elements that are exclusive to the original image as deleted elements, count the deleted elements of all reference data, mark the number of marks corresponding to the deleted elements as the elimination value of the deleted elements, mark the L2 deleted elements with the largest elimination value as the deletion execution elements, and eliminate the deletion execution elements in the drawing task; Step K3: Compare the original image in the reference data with the generated image: mark the image elements that are unique to the generated image as added elements, count the added elements of all reference data, mark the number of markings corresponding to the added elements as the added values ​​of the added elements, mark the L2 added elements with the largest added values ​​as added execution elements, and add the added execution elements to the drawing task.

2. According to claim 1, an AI-based intelligent mapping system is characterized in that: The process of obtaining the characteristic parameter range includes: obtaining the original rich low value FFmin and the original rich high value FFmax through the formula FFmin=p1*FFy and FFmax=p2*FFy, where p1 and p2 are both proportional coefficients, and 0.85≤p1≤0.95; 1.05≤p2≤1.15; the rich range is composed of the original rich low value FFmin and the original rich high value FFmax; obtaining the original smooth low value PHmin and the original smooth high value PHmax through the formula PHmin=p3*PHy and PHmax=p4*PHy, where p3 and p4 are both proportional coefficients, And 0.85≤p3≤0.95; 1.05≤p4≤1.15; the smoothing range is composed of the original smoothing low value PHmin and the original smoothing high value PHmax; the original color low value YSmin and the original color high value YSmax are obtained by the formulas YSmin=p5*YSy and YSmax=p6*YSy, where p5 and p6 are both proportional coefficients, and 0.85≤p5≤0.95; 1.05≤p6≤1.15; the color range is composed of the original color low value YSmin and the original color high value YSmax; the feature parameter range is composed of the rich range, smoothing range and color range.

Images