CN-122020811-A - Method, apparatus, device, medium and program product for generating vector structure data

CN122020811ACN 122020811 ACN122020811 ACN 122020811ACN-122020811-A

Abstract

The present application relates to a method, apparatus, device, medium and program product for generating vector structure data. The method comprises the steps of obtaining a target house type graph, carrying out graph recognition processing on the target house type graph to obtain graph recognition data, carrying out text recognition processing on the target house type graph to obtain text recognition data, and generating vector structure data corresponding to the target house type graph according to the graph recognition data and the text recognition data. By adopting the method, the accuracy of the vector structure data can be improved.

Inventors

FAN GANG
LI YUNZHU
Lin Langqing

Assignees

厦门阿匹斯智能机器人有限公司

Dates

Publication Date: 20260512
Application Date: 20260324

Claims (10)

1. A method for generating vector structure data, the method comprising: obtaining a target house type diagram; performing pattern recognition processing on the target house pattern to obtain pattern recognition data, and Performing text recognition processing on the target house type graph to obtain text recognition data; And generating vector structure data corresponding to the target house type graph according to the pattern recognition data and the text recognition data.
2. The method of claim 1, wherein the pattern recognition data includes topology data between pattern elements in the target house type graph, the text recognition data includes text content of different target texts in the target house type graph, and the generating vector structure data corresponding to the target house type graph according to the pattern recognition data and the text recognition data includes: Determining a matching relationship between the graphic element and the target text; Based on the matching relation, generating vector structure data corresponding to the target house type graph according to the topological structure data among the graphic elements and the text content of the different target texts.
3. The method of claim 2, wherein the determining a matching relationship between the graphical element and the target text comprises at least one of: For each target text, determining a target graphic element matched with the target text from the graphic elements according to the text content of the target text; For each target text, determining a target graphic element matched with the target text from the graphic elements according to the text position of the target text in the text identification data; for each target text, determining a target area associated with the target text according to the text position of the target text in the text recognition data, wherein the target area is an area formed by at least one target graphic element.
4. The method of claim 2, wherein the text content includes annotation data, and the generating, based on the matching relationship, vector structure data corresponding to the target house type graph according to topology structure data between the graphic elements and text content of the different target texts, includes at least one of: Based on the matching relation, determining a first geometric dimension corresponding to a graphic element matched with the target text according to the labeling data in the target text, and generating vector structure data corresponding to the target house type graph according to topological structure data among the graphic elements and the first geometric dimension of different graphic elements; Based on the matching relation, determining a first geometric dimension corresponding to a graphic element matched with the target text according to marking data in the target text, acquiring a second geometric dimension corresponding to different graphic elements in the graphic identification data, and generating vector structure data corresponding to the target house type graph according to topological structure data among the graphic elements, the first geometric dimension different from the graphic elements and the second geometric dimension different from the graphic elements.
5. The method according to any one of claims 1-4, wherein the text recognition data includes text content and/or text location, and the corresponding text recognition processing is performed on the target floor plan to obtain text recognition data, including at least one of: Determining at least one text area in the target house type graph, and identifying text content in each text area; and determining at least one text area in the target house type graph, and determining the text position of the text area in the target house type graph.
6. The method according to any one of claims 1-4, wherein the pattern recognition data includes topology data among graphic elements in the target house type graph, and the performing pattern recognition processing on the target house type graph to obtain the pattern recognition data includes: Extracting image features in the target house type graph; According to the image characteristics, carrying out semantic segmentation processing on the target house type graph to obtain semantic segmentation data of the target house type graph, wherein the semantic segmentation data comprises segmentation results of different graphic elements; And determining topological structure data among the graphic elements according to the segmentation results of different graphic elements.
7. A device for generating vector structure data, the device comprising: the acquisition module is used for acquiring the target house type graph; a first recognition module for performing pattern recognition processing on the target house pattern to obtain pattern recognition data, and The second recognition module is used for carrying out text recognition processing on the target house type graph to obtain text recognition data; and the generating module is used for generating vector structure data corresponding to the target house type graph according to the pattern recognition data and the text recognition data.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

Description

Method, apparatus, device, medium and program product for generating vector structure data Technical Field The present application relates to the field of data processing technologies, and in particular, to a method, an apparatus, a device, a medium, and a program product for generating vector structure data. Background The vector structure data can describe the positions and the sizes of elements such as walls, doors and windows more accurately, and data support is provided for subsequent processes such as area calculation, space analysis and space layout. The editable and computable vector structure data based on the house type drawing in paper or image form is called as an important work in the fields of architectural design, interior decoration and the like. However, the traditional vector structure data generation mode mainly relies on manual drawing, namely corresponding vector structure data is required to be manually drawn according to a house type drawing, so that deviation between the vector data and an original drawing is easily caused by manual interpretation errors or misoperation, and the accuracy of the vector structure data is affected. Disclosure of Invention In view of the foregoing, it is desirable to provide a method, apparatus, device, medium, and program product for generating vector structure data, so as to improve the accuracy of the vector structure data. In a first aspect, the present application provides a method for generating vector structure data, including: obtaining a target house type diagram; performing pattern recognition processing on the target house type graph to obtain pattern recognition data, and Performing text recognition processing on the target house type graph to obtain text recognition data; And generating vector structure data corresponding to the target house type graph according to the graph identification data and the text identification data. In one embodiment, the pattern recognition data comprises topological structure data among pattern elements in the target house type graph, the text recognition data comprises text contents of different target texts in the target house type graph, and correspondingly, generating vector structure data corresponding to the target house type graph according to the pattern recognition data and the text recognition data comprises determining a matching relation between the pattern elements and the target texts, and generating vector structure data corresponding to the target house type graph according to the topological structure data among the pattern elements and the text contents of the different target texts based on the matching relation. In one embodiment, the matching relation between the graphic element target texts is determined, and the matching relation comprises at least one of determining target graphic elements matched with the target texts from graphic elements according to text contents of the target texts, determining target graphic elements matched with the target texts from graphic elements according to text positions of the target texts in text recognition data for each target text, and determining target areas associated with the target texts according to text positions of the target texts in the text recognition data for each target text, wherein the target areas are areas formed by at least one target graphic element. In one embodiment, the text content comprises marking data, corresponding vector structure data corresponding to a target house type graph is generated according to topological structure data among graphic elements and text content of different target texts based on a matching relation, the method comprises at least one of determining first geometric dimensions corresponding to graphic elements matched with the target text based on the matching relation according to the marking data in the target text, generating vector structure data corresponding to the target house type graph according to the topological structure data among the graphic elements and the first geometric dimensions of different graphic elements, determining first geometric dimensions corresponding to graphic elements matched with the target text based on the matching relation according to the marking data in the target text, obtaining second geometric dimensions corresponding to different graphic elements in the graphic identification data, and generating vector structure data corresponding to the target house type graph according to the topological structure data among the graphic elements, the first geometric dimensions of different graphic elements and the second geometric dimensions of different graphic elements. In one embodiment, the text recognition data comprises text content and/or text position, and the corresponding text recognition processing is carried out on the target house type graph to obtain the text recognition data, wherein the text recognition data comprises at least one of determining at least one