Robot drawing method and system based on voice interaction

Abstract

The invention is suitable for the technical field of robot painting, and provides a robot painting method and system based on voice interaction, and the method comprises the following steps: collecting a voice description instruction, and determining keyword information according to the voice description instruction, and the keyword information comprises a painting main body, a main body relation, a style and a color; inputting the keyword information into an AI drawing model, and automatically generating a drawing picture; extracting element information and contour information in the drawing picture, performing simplification and abstraction processing on the element information and the contour information, and converting the element information and the contour information into a simple pattern; acquiring a voice modification instruction, and modifying and editing the simple pattern based on the voice modification instruction; and converting the simplified pattern into vector diagram data based on a vector graphic algorithm, so that the robot executes a drawing action according to the vector diagram data. According to the method, the drawing picture is converted into the simple pattern suitable for being drawn by the robot, the drawing artistry is reserved, the execution complexity of the robot is reduced, and then the hardware cost of the drawing robot is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of robot painting, and in particular to a robot painting method and system based on voice interaction. Background Art

[0002] With the rapid development of artificial intelligence and robotics, intelligent painting robots have gradually become a research hotspot in the field of artistic creation and human-computer interaction. Traditional painting robots usually rely on pre-programmed instructions or manually input graphic data to perform painting tasks. This method lacks flexibility and interactivity and is difficult to meet users' needs for personalized creation. In recent years, technologies based on voice interaction and AI painting models have provided new possibilities for intelligent painting robots, enabling users to generate paintings through natural language descriptions. However, existing intelligent painting robots still have the following problems: limited ability to parse voice descriptions, difficulty in accurately extracting key information in painting tasks, resulting in the generated paintings not being consistent with user intentions; most painting models only support the generation of paintings from voice descriptions, lack of real-time modification functions for generated works, and users cannot adjust the details of paintings through voice commands; lack of simplification and abstract processing of painting elements, high pattern complexity, resulting in high hardware requirements for painting robots; in addition, traditional methods use bitmap data directly for robot painting, lack of support for vector graphics data, resulting in low painting accuracy and efficiency. Therefore, it is necessary to provide a robot painting method and system based on voice interaction to solve the above problems. Summary of the invention

[0003] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a robot painting method and system based on voice interaction to solve the problems existing in the above-mentioned background technology.

[0004] The present invention is implemented as follows: a robot painting method based on voice interaction, the method comprising the following steps: Collecting voice description instructions, and determining keyword information according to the voice description instructions, wherein the keyword information includes painting subject, subject relationship, style and color; Input keyword information into the AI ​​painting model to automatically generate painting images; Extract element information and contour information from the painting image, simplify and abstract the element information and contour information, and convert them into simple strokes; Collect voice modification instructions, and modify and edit the stick figure pattern based on the voice modification instructions; The stick figure pattern is converted into vector graphic data based on the vector graphic algorithm, so that the robot can perform drawing actions according to the vector graphic data.

[0005] As a further solution of the present invention: the step of determining keyword information according to the voice description instruction specifically includes: Preprocessing the voice description instructions, including denoising, filtering and volume normalization, and converting the voice description instructions into text information based on the speech recognition algorithm; Perform word segmentation and stop word removal on the identified text information to identify individual words or phrases; Based on named entity recognition technology, entities related to paintings are extracted to obtain the painting subject, style and color; A dual-subject phrase is extracted from the text information to obtain a subject relationship, wherein the dual-subject phrase includes a qualifier and at least two phrases, and the qualifier is a verb or a directional word.

[0006] As a further solution of the present invention: the step of extracting element information and contour information from the painting image specifically includes: Grayscale and denoise the painting image, identify the edges in the image based on the Canny edge detection algorithm, and generate an edge image; The image is divided into different regions based on the region segmentation algorithm, each region corresponds to an element; For each segmented region, the contour point set of the region is extracted to form contour information, and the original color and shape features of each region are recorded as element information.

[0007] As a further solution of the present invention: the step of simplifying and abstracting the element information and the outline information to convert them into a simple pattern specifically includes: The contour information is simplified based on the Douglas-Peucker algorithm, and the farthest point of the offline segment is recursively deleted until the simplification threshold is met; Use polygonal approximation algorithm to abstract shape features in element information into simple geometric shapes; Match the original color in each element with the robot supported color collection to determine the available color corresponding to each element; Determine whether the elements corresponding to the same available color are adjacent in the drawing picture. If they are adjacent, determine the color difference between the corresponding original colors. When the color difference is greater than the set difference, modify one of the available colors.

[0008] As a further solution of the present invention: the step of collecting voice modification instructions and modifying and editing the stick figure pattern based on the voice modification instructions specifically includes: Identify each contour line and each color block in the stick figure pattern, number each contour line and each color block, and display the numbered stick figure pattern; A voice modification instruction is collected, wherein the voice modification instruction includes a number and a corresponding modification suggestion, and the stick figure pattern is modified and edited based on the voice modification instruction.

[0009] As a further solution of the present invention: the step of converting the stick figure pattern into vector graphic data based on the vector graphics algorithm specifically includes: The outline of the stick figure is fitted by a Bezier curve or a spline curve to generate a smooth vector outline and obtain a vector graphic; According to the color information of the element, a color code is assigned to each area of ​​the vector graphic to obtain the fill information, and all the vector outline and fill information are integrated into the vector graphic data.

[0010] Another object of the present invention is to provide a robot painting system based on voice interaction, the system comprising: A keyword information determination module, used to collect voice description instructions and determine keyword information according to the voice description instructions, wherein the keyword information includes painting subject, subject relationship, style and color; A painting image generation module is used to input keyword information into the AI ​​painting model to automatically generate painting images; A simple pattern generation module is used to extract element information and contour information from the painting image, simplify and abstract the element information and contour information, and convert them into simple patterns; A stick figure pattern modification module is used to collect voice modification instructions and modify and edit the stick figure pattern based on the voice modification instructions; The vector graphics data acquisition module is used to convert the stick figure pattern into vector graphics data based on the vector graphics algorithm, so that the robot can perform the drawing action according to the vector graphics data.

[0011] As a further solution of the present invention: the keyword information determination module includes: A speech preprocessing unit, used for preprocessing the speech description instruction, the preprocessing includes denoising, filtering and volume normalization, and converting the speech description instruction into text information based on the speech recognition algorithm; A vocabulary and phrase recognition unit is used to perform word segmentation and stop word removal on the recognized text information to recognize a single word or phrase; A painting entity extraction unit is used to extract entities related to the painting based on named entity recognition technology to obtain the painting subject, style and color; The subject relationship determination unit is used to extract dual-subject phrases in text information to obtain subject relationships, wherein the dual-subject phrases include a qualifier and at least two phrases, and the qualifier is a verb or a directional word.

[0012] As a further solution of the present invention: the stick figure pattern generation module comprises: The image edge recognition unit is used to grayscale and denoise the painting image, recognize the edge in the image based on the Canny edge detection algorithm, and generate an edge image; An image region segmentation unit, used for segmenting the image into different regions based on a region segmentation algorithm, each region corresponding to one element; The contour element determination unit is used to extract the contour point set of each segmented area, form contour information, and record the original color and shape characteristics of each area as element information.

[0013] As a further solution of the present invention: the stick figure pattern generation module further includes: A contour information simplification unit is used to simplify the contour information based on the Douglas-Peucker algorithm, and recursively delete the farthest point of the offline segment until the simplification threshold is met; A shape feature abstraction unit is used to abstract the shape features in the element information into simple geometric shapes using a polygon approximation algorithm; An available color determination unit, used to match the original color in each element with the robot supported color collection to determine the available color corresponding to each element; The available color modification unit is used to determine whether the elements corresponding to the same available color are adjacent in the painting picture. When they are adjacent, the color difference between the corresponding original colors is determined. When the color difference is greater than the set difference, one of the available colors is modified.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The present invention will determine keyword information according to the voice description instruction, and the keyword information includes the painting subject, subject relationship, style and color, to ensure that the generated painting works meet the user's intention. It will also extract the element information and contour information in the painting picture, simplify and abstract the element information and contour information, and convert the painting picture into a simple pattern suitable for robot drawing, which not only retains the artistry of the painting, but also reduces the complexity of the robot's execution, thereby reducing the hardware cost of the painting robot and making it more applicable. In addition, the present invention will also modify and edit the simple pattern according to the voice modification instruction to enhance the interactive experience and creative flexibility. It will also convert the simple pattern into vector data based on the vector graphics algorithm to ensure that the robot can efficiently and accurately perform painting actions and improve the quality and efficiency of painting. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The flowchart of a robot painting method based on voice interaction is shown in FIG.

[0016] Figure 2It is a flowchart for determining keyword information in a robot painting method based on voice interaction.

[0017] Figure 3 It is a flowchart for extracting element information and contour information in a robot painting method based on voice interaction.

[0018] Figure 4 It is a flowchart for converting into a simple stick figure pattern in a robot painting method based on voice interaction.

[0019] Figure 5 It is a flowchart for collecting voice modification instructions in a robot painting method based on voice interaction.

[0020] Figure 6 It is a flowchart for converting into vector graphic data in a robot painting method based on voice interaction.

[0021] Figure 7 It is a schematic structural diagram of a robot painting system based on voice interaction. Detailed implementation manners

[0022] To make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0023] The following describes the specific implementation of the present invention in detail with reference to specific embodiments.

[0024] As Figure 1 shown, an embodiment of the present invention provides a robot painting method based on voice interaction, and the method includes the following steps: S100, collect voice description instructions, and determine keyword information according to the voice description instructions, where the keyword information includes a painting subject, a subject relationship, a style, and a color; S200, input the keyword information into an AI painting model to automatically generate a painting picture; S300, extract element information and contour information in the painting picture, perform simplification and abstraction processing on the element information and contour information, and convert it into a simple stick figure pattern; S400, collect voice modification instructions, and modify and edit the simple stick figure pattern based on the voice modification instructions; S500, convert the simple stick figure pattern into vector graphic data based on a vector graphics algorithm, so that the robot performs a painting action according to the vector graphic data.

[0025] It should be noted that existing intelligent painting robots still have the following problems: limited ability to parse voice descriptions, difficulty in accurately extracting key information in painting tasks, resulting in the generated paintings not being consistent with user intent; most painting models only support generating paintings from voice descriptions, lacking the function of real-time modification of generated works, and users cannot adjust painting details through voice commands; lack of simplification and abstract processing of painting elements, high complexity of patterns, resulting in high hardware requirements for painting robots. The embodiments of the present invention are intended to solve the above problems.

[0026] In the embodiment of the present invention, first, the user needs to input a voice description instruction to describe what he wants to draw. The embodiment of the present invention will determine the keyword information according to the voice description instruction, and the keyword information includes the painting subject, subject relationship, style and color to ensure that the generated painting works meet the user's intention. Then the keyword information is input into the AI ​​painting model to automatically generate a painting picture. Common AI painting models include Wenxin Yiyan, DALL·E series, CLIP-Guided Diffusion, etc. Then the element information and contour information in the painting picture will be extracted, and the element information and contour information will be simplified and abstracted, and the painting picture will be converted into a simple pattern suitable for robot drawing, which not only retains the artistry of the painting, but also reduces the complexity of the robot execution, thereby reducing the hardware cost of the painting robot and making it more applicable. After the simple pattern is generated, it will be displayed on the display. The user can input a voice modification instruction according to the needs. The embodiment of the present invention will modify and edit the simple pattern according to the voice modification instruction to enhance the interactive experience and creative flexibility. Finally, the simple pattern will be converted into vector data based on the vector graphics algorithm to ensure that the robot can efficiently and accurately perform painting actions and improve the quality and efficiency of painting.

[0027] like Figure 2 As shown, as a preferred embodiment of the present invention, the step of determining keyword information according to the voice description instruction specifically includes: S101, preprocessing the voice description instruction, the preprocessing including denoising, filtering and volume normalization, and converting the voice description instruction into text information based on a speech recognition algorithm; S102, performing word segmentation and stop word removal processing on the recognized text information to recognize a single word or phrase; S103, extracting entities related to the painting based on named entity recognition technology to obtain the subject, style and color of the painting; S104, extracting dual-subject phrases in the text information to obtain subject relations, wherein the dual-subject phrases include a qualifier and at least two phrases, and the qualifier is a verb or a directional word.

[0028] In the embodiment of the present invention, a high-fidelity microphone is used to collect the user's voice instructions, and the voice description instructions are preprocessed, including denoising, filtering and volume normalization to improve the accuracy of voice recognition, and the voice description instructions are converted into text information based on a voice recognition algorithm (such as an ASR model based on deep learning, such as DeepSpeech or Wav2Vec). Then, the recognized text information is segmented and stop words are removed to identify single words or phrases, and then named entity recognition (NER) technology is used to extract entities related to the painting to obtain the subject, style and color of the painting, and dual-subject phrases in the text information are extracted to obtain the subject relationship, the dual-subject phrases include a qualifier and at least two phrases, the qualifier is a verb or a directional word, for example, the dual-subject phrases are: a bird is on the roof, a boy is playing football, etc.

[0029] like Figure 3 As shown, as a preferred embodiment of the present invention, the step of extracting element information and contour information from the painting image specifically includes: S301, graying and denoising the painting image, identifying the edge in the image based on the Canny edge detection algorithm, and generating an edge image; S302, segmenting the image into different regions based on a region segmentation algorithm, each region corresponding to an element; S303, for each segmented region, extract the contour point set of the region to form contour information, and record the original color and shape features of each region as element information.

[0030] In the embodiment of the present invention, in order to better identify the contour edge, the painting image is grayed and denoised, and the edge in the image is identified based on the Canny edge detection algorithm to obtain an edge image. Then, the image is segmented into different regions according to a region segmentation algorithm (such as a flood fill algorithm), and each region corresponds to an element. Furthermore, the regions can be further subdivided or merged according to features such as color and texture to optimize element recognition. Finally, for each segmented region, the contour point set of the region is extracted, and the original color and shape features of each region are recorded as element information.

[0031] like Figure 4 As shown, as a preferred embodiment of the present invention, the step of simplifying and abstracting the element information and the outline information and converting them into a simple pattern specifically includes: S304, simplifying the contour information based on the Douglas-Peucker algorithm, recursively deleting the farthest point of the offline segment until a simplification threshold is met; S305, using a polygonal approximation algorithm to abstract shape features in the element information into simple geometric shapes; S306, matching the original color in each element with the robot supported color collection to determine the available color corresponding to each element; S307, determining whether elements corresponding to the same available color are adjacent in the drawing picture, and if so, determining the color difference between the corresponding original colors, and when the color difference is greater than the set difference, modifying one of the available colors.

[0032] In the embodiment of the present invention, when performing simplification and abstraction processing, the Douglas-Peucker algorithm is applied to simplify the contour information, and the farthest point of the offline segment is recursively deleted until the simplification threshold is met. The simplification threshold needs to be set in advance to reduce unnecessary detail points and retain the main shape features. Then, a polygon approximation algorithm (such as a convex hull algorithm or a minimum circumscribed rectangle algorithm) is used to abstract the complex shape features in the element information into simple geometric shapes. Next, the original color in each element is matched with the robot support color collection, which refers to the collection of all colors that the painting robot can draw, and the most similar available color in the color collection is determined to be the available color corresponding to each element. In addition, in order to improve the presentation effect, it is also determined whether the elements corresponding to the same available color are adjacent in the painting picture. When adjacent, the color difference value between the corresponding original colors is further determined. When the color difference value is greater than the set difference value, the difference is set to a fixed value, and one of the available colors needs to be modified so that the available colors are different and the sense of hierarchy is highlighted.

[0033] like Figure 5 As shown, as a preferred embodiment of the present invention, the step of collecting voice modification instructions and modifying and editing the stick figure pattern based on the voice modification instructions specifically includes: S401, identifying each contour line and each color block in the stick figure pattern, numbering each contour line and each color block, and displaying the numbered stick figure pattern; S402, collecting voice modification instructions, where the voice modification instructions include numbers and corresponding modification suggestions, and modifying and editing the stick figure pattern based on the voice modification instructions.

[0034] In the embodiment of the present invention, in order to facilitate the user to perform voice modification and improve the accuracy of the modification, the embodiment of the present invention automatically identifies each contour line and each color block in the stick figure pattern, and numbers each contour line and each color block, and then displays the numbered stick figure pattern on the display. In this way, the user can intuitively speak the modification voice, and the embodiment of the present invention will collect the voice modification instruction, which includes the number and the corresponding modification opinion, such as making the contour line 012 more rounded and making the color of the color block 021 red.

[0035] like Figure 6As shown, as a preferred embodiment of the present invention, the step of converting the stick figure pattern into vector graphic data based on the vector graphics algorithm specifically includes: S501, fitting the outline of the stick figure pattern by a Bezier curve or a spline curve to generate a smooth vector outline and obtain a vector graphic; S502, assigning a color code to each region of the vector graphic according to the color information of the element, obtaining fill information, and integrating all vector outlines and fill information into vector graphic data.

[0036] In an embodiment of the present invention, in order to obtain vector graphics data, firstly, the outline in the stick figure is fitted by a Bezier curve or a spline curve to generate a smooth vector outline, and the control points of the curve are adjusted to ensure that the shape of the curve is consistent with the original outline to obtain a vector graphic. Then, according to the color information of the element, a color code is assigned to each area of ​​the vector graphic to obtain fill information, and all vector outlines and fill information are integrated into vector graphics data. Specifically, all vector outlines and fill information are integrated into the data structure of a vector graphics file (such as SVG format), and the SVG file contains path commands (such as M, L, C, etc.) and attributes (such as stroke, fill, etc.) for describing the geometric shape and style of the graphics. Finally, the vector graphics data is output for use by the robot when performing the painting action.

[0037] like Figure 7 As shown, an embodiment of the present invention further provides a robot painting system based on voice interaction, the system comprising: The keyword information determination module 100 is used to collect voice description instructions and determine keyword information according to the voice description instructions, wherein the keyword information includes painting subject, subject relationship, style and color; A painting picture generation module 200 is used to input keyword information into the AI ​​painting model to automatically generate a painting picture; A simple pattern generation module 300 is used to extract element information and contour information from a painting image, simplify and abstract the element information and contour information, and convert them into a simple pattern; A stick figure pattern modification module 400 is used to collect voice modification instructions and modify and edit the stick figure pattern based on the voice modification instructions; The vector graphics data acquisition module 500 is used to convert the stick figure pattern into vector graphics data based on a vector graphics algorithm, so that the robot can perform a drawing action according to the vector graphics data.

[0038] As a preferred embodiment of the present invention, the keyword information determination module 100 includes: A speech preprocessing unit, used for preprocessing the speech description instruction, the preprocessing includes denoising, filtering and volume normalization, and converting the speech description instruction into text information based on the speech recognition algorithm; A vocabulary and phrase recognition unit is used to perform word segmentation and stop word removal on the recognized text information to recognize a single word or phrase; A painting entity extraction unit is used to extract entities related to the painting based on named entity recognition technology to obtain the painting subject, style and color; The subject relationship determination unit is used to extract dual-subject phrases in text information to obtain subject relationships, wherein the dual-subject phrases include a qualifier and at least two phrases, and the qualifier is a verb or a directional word.

[0039] As a preferred embodiment of the present invention, the stick figure pattern generating module 300 includes: The image edge recognition unit is used to grayscale and denoise the painting image, recognize the edge in the image based on the Canny edge detection algorithm, and generate an edge image; An image region segmentation unit, used for segmenting the image into different regions based on a region segmentation algorithm, each region corresponding to one element; The contour element determination unit is used to extract the contour point set of each segmented area, form contour information, and record the original color and shape characteristics of each area as element information.

[0040] As a preferred embodiment of the present invention, the stick figure pattern generating module 300 further includes: A contour information simplification unit is used to simplify the contour information based on the Douglas-Peucker algorithm, and recursively delete the farthest point of the offline segment until the simplification threshold is met; A shape feature abstraction unit is used to abstract the shape features in the element information into simple geometric shapes using a polygon approximation algorithm; An available color determination unit, used to match the original color in each element with the robot supported color collection to determine the available color corresponding to each element; The available color modification unit is used to determine whether the elements corresponding to the same available color are adjacent in the painting picture. When they are adjacent, the color difference between the corresponding original colors is determined. When the color difference is greater than the set difference, one of the available colors is modified.

[0041] The above only describes in detail the preferred embodiments of the present invention, which is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0042] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0043] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0044] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the disclosure in the specification and examples. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

Claims

1. A robot painting method based on voice interaction, characterized in that: The method comprises the following steps: Collecting voice description instructions, and determining keyword information according to the voice description instructions, wherein the keyword information includes painting subject, subject relationship, style and color; Input keyword information into the AI ​​painting model to automatically generate painting images; Extract element information and contour information from the painting image, simplify and abstract the element information and contour information, and convert them into simple strokes; Collect voice modification instructions, and modify and edit the stick figure pattern based on the voice modification instructions; The stick figure pattern is converted into vector graphic data based on the vector graphic algorithm, so that the robot can perform drawing actions according to the vector graphic data.

2. The robot painting method based on voice interaction according to claim 1, characterized in that: The step of determining keyword information according to the voice description instruction specifically includes: Preprocessing the voice description instructions, including denoising, filtering and volume normalization, and converting the voice description instructions into text information based on the speech recognition algorithm; Perform word segmentation and stop word removal on the identified text information to identify individual words or phrases; Based on named entity recognition technology, entities related to paintings are extracted to obtain the painting subject, style and color; A dual-subject phrase is extracted from the text information to obtain a subject relationship, wherein the dual-subject phrase includes a qualifier and at least two phrases, and the qualifier is a verb or a directional word.

3. The robot painting method based on voice interaction according to claim 1, characterized in that: The step of extracting element information and contour information from the painting image specifically includes: Grayscale and denoise the painting image, identify the edges in the image based on the Canny edge detection algorithm, and generate an edge image; The image is divided into different regions based on the region segmentation algorithm, each region corresponds to an element; For each segmented region, the contour point set of the region is extracted to form contour information, and the original color and shape features of each region are recorded as element information.

4. The robot painting method based on voice interaction according to claim 3 is characterized in that: The step of simplifying and abstracting the element information and the outline information to convert them into a simple pattern specifically includes: The contour information is simplified based on the Douglas-Peucker algorithm, and the farthest point of the offline segment is recursively deleted until the simplification threshold is met; Use polygonal approximation algorithm to abstract shape features in element information into simple geometric shapes; Match the original color in each element with the robot supported color collection to determine the available color corresponding to each element; Determine whether the elements corresponding to the same available color are adjacent in the drawing picture. If they are adjacent, determine the color difference between the corresponding original colors. When the color difference is greater than the set difference, modify one of the available colors.

5. The robot painting method based on voice interaction according to claim 1, characterized in that: The step of collecting voice modification instructions and modifying and editing the stick figure pattern based on the voice modification instructions specifically includes: Identify each contour line and each color block in the stick figure pattern, number each contour line and each color block, and display the numbered stick figure pattern; A voice modification instruction is collected, wherein the voice modification instruction includes a number and a corresponding modification suggestion, and the stick figure pattern is modified and edited based on the voice modification instruction.

6. The robot painting method based on voice interaction according to claim 4, characterized in that: The step of converting the stick figure pattern into vector graphics data based on the vector graphics algorithm specifically includes: The outline of the stick figure is fitted by a Bezier curve or a spline curve to generate a smooth vector outline and obtain a vector graphic; According to the color information of the element, a color code is assigned to each area of ​​the vector graphic to obtain the fill information, and all the vector outline and fill information are integrated into the vector graphic data.

7. A robot painting system based on voice interaction, characterized in that: The system comprises: A keyword information determination module, used to collect voice description instructions and determine keyword information according to the voice description instructions, wherein the keyword information includes painting subject, subject relationship, style and color; A painting image generation module is used to input keyword information into the AI ​​painting model to automatically generate painting images; A simple pattern generation module is used to extract element information and contour information from the painting image, simplify and abstract the element information and contour information, and convert them into simple patterns; A stick figure pattern modification module is used to collect voice modification instructions and modify and edit the stick figure pattern based on the voice modification instructions; The vector graphics data acquisition module is used to convert the stick figure pattern into vector graphics data based on the vector graphics algorithm, so that the robot can perform the drawing action according to the vector graphics data.

8. The robot painting system based on voice interaction according to claim 7, characterized in that: The keyword information determination module includes: A speech preprocessing unit, used for preprocessing the speech description instruction, the preprocessing includes denoising, filtering and volume normalization, and converting the speech description instruction into text information based on the speech recognition algorithm; A vocabulary and phrase recognition unit is used to perform word segmentation and stop word removal on the recognized text information to recognize a single word or phrase; A painting entity extraction unit is used to extract entities related to the painting based on named entity recognition technology to obtain the painting subject, style and color; The subject relationship determination unit is used to extract dual-subject phrases in text information to obtain subject relationships, wherein the dual-subject phrases include a qualifier and at least two phrases, and the qualifier is a verb or a directional word.

9. The robot painting system based on voice interaction according to claim 7, characterized in that: The stick figure pattern generating module comprises: The image edge recognition unit is used to grayscale and denoise the painting image, recognize the edge in the image based on the Canny edge detection algorithm, and generate an edge image; An image region segmentation unit, used for segmenting the image into different regions based on a region segmentation algorithm, each region corresponding to one element; The contour element determination unit is used to extract the contour point set of each segmented area, form contour information, and record the original color and shape characteristics of each area as element information.

10. The robot painting system based on voice interaction according to claim 9, characterized in that: The stick figure pattern generating module also includes: A contour information simplification unit is used to simplify the contour information based on the Douglas-Peucker algorithm, and recursively delete the farthest point of the offline segment until the simplification threshold is met; A shape feature abstraction unit is used to abstract the shape features in the element information into simple geometric shapes using a polygon approximation algorithm; An available color determination unit, used to match the original color in each element with the robot supported color collection to determine the available color corresponding to each element; The available color modification unit is used to determine whether the elements corresponding to the same available color are adjacent in the painting picture. When they are adjacent, the color difference between the corresponding original colors is determined. When the color difference is greater than the set difference, one of the available colors is modified.

Images