A structural column attribute recognition method based on traditional image processing combined with deep learning

Abstract

The present invention discloses a method for identifying the properties of structural professional columns based on traditional image processing combined with deep learning, comprising the following steps: firstly preparing data; parsing drawings to obtain a series of primitives; obtaining all primitive lines in the column layer; using traditional image processing methods to find the space of all columns; using traditional image processing methods to find text information with "Z" near the column; using deep learning methods to obtain the position of the column property table, obtain the position column number and property information; and obtain the properties of the column by matching the column number with the column number in the table. The present invention solves the problem of accurate identification of components in CAD architectural design drawings by using traditional image processing methods, reduces the excessive dependence of component identification on layers, and by using deep learning target detection methods to identify column property tables, it can effectively solve the confusion of table styles and positions, and improve the accuracy and speed of obtaining table positions through logical judgment.

Description

Technical Field

[0001] The present invention relates to the field of column attribute recognition in a flat construction drawing of a structural professional residence or basement wall column, and in particular to a column attribute recognition method based on a target detection technology combined with traditional image processing and deep learning. Background Art

[0002] CAD construction drawings are drawings that use AutoCAD software to show the overall layout of the project, the building's exterior shape, interior layout, structural structure, interior and exterior decoration, material methods, equipment, and construction. CAD construction drawings are complete, accurate, and specific. They are the basis for engineering construction, construction drawing budgets, and construction organization design. They are also important technical documents for technical management. Before construction, the construction drawings need to be carefully reviewed before entering the construction phase. The purpose is to ensure the smooth progress of the construction and avoid the impact of errors in the drawings on the use phase after the completion of the construction.

[0003] The existing identification of structural professional columns is mainly based on traditional methods such as layers, manual annotation, and a large number of logical stacking. This method is time-consuming and labor-intensive, and in the process of merging through business logic, due to the unfixed style and position of attribute text and attribute table, there are two main defects:

[0004] 1. Identifying column properties through layers in drawings is heavily dependent on drawing design specifications.

[0005] 2. The accuracy of attribute table obtained through logic is low, resulting in low attribute accuracy. Summary of the invention

[0006] In response to the above problems, the present invention provides a method for identifying structural professional column attributes based on traditional image processing combined with deep learning.

[0007] To solve the above technical problems, the present invention is implemented by the following technical solutions: The present invention provides a method for identifying the properties of structural professional columns based on traditional image processing combined with deep learning, the method comprising the following steps:

[0008] S1. Prepare data first;

[0009] S2, parsing the drawing to obtain a series of graphic elements;

[0010] S3, get all the graphic element lines in the column layer, and draw these graphic element lines on a blank base map;

[0011] S4, using traditional image processing methods to find the space of all columns;

[0012] S5. Using traditional image processing methods, find the text information with "Z" near the column;

[0013] S6. Use deep learning methods to obtain the column attribute table location, obtain the location column number and attribute information;

[0014] S7. Obtain the attributes of the column by matching the column number with the column number in the table.

[0015] Preferably, the data preparation in step S1 specifically includes: 30 residential or basement wall and column flat construction drawings.

[0016] Preferably, the graphic elements in step S2 include visible basic elements constituting graphics, which correspond to visible entities on the drawing interface.

[0017] Preferably, in step S4, the spaces of all columns are filtered out through the hatch component bbox range to filter out interfering space columns.

[0018] Preferably, in step S5, the column number can be obtained by filtering through regular rules.

[0019] Preferably, the table detection model trained with Ret i naNet used in step S6.

[0020] It can be seen from the above description of the present invention that, compared with the prior art, the present invention has the following advantages:

[0021] The present invention discloses a structural professional column attribute recognition method based on traditional image processing combined with deep learning. By applying traditional image processing methods, the problem of accurate identification of components in CAD architectural design drawings is solved, and excessive reliance on layers for component identification is reduced. In addition, by using deep learning target detection methods to identify column attribute tables, the confusion of table styles and positions can be effectively resolved, and the accuracy and speed of obtaining table positions through logical judgment can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a flow chart of the steps of the present invention;

[0023] Figure 2 It is a schematic diagram of the CAD drawing analysis of the flat construction drawing of the wall column of a house or basement of the present invention;

[0024] Figure 3 It is a CAD plane schematic diagram of column space acquisition of the present invention;

[0025] Figure 4 It is a CAD plan schematic diagram of the column space filtering of the present invention;

[0026] Figure 5It is a CAD plan schematic diagram of column properties detected by deep learning in the present invention;

[0027] Figure 6 It is a CAD plane schematic diagram of the present invention that obtains attributes through number matching;

[0028] Figure 7 It is a schematic diagram of the integration of column attribute information of the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] like Figure 1-7 As shown, the present invention provides a method for identifying the properties of structural professional columns based on traditional image processing combined with deep learning, and the method comprises the following steps:

[0031] S1. Prepare data first;

[0032] S2, parsing the drawing to obtain a series of graphic elements;

[0033] S3, get all the graphic element lines in the column layer, and draw these graphic element lines on a blank base map;

[0034] S4, using traditional image processing methods to find the space of all columns;

[0035] S5. Using traditional image processing methods, find the text information with "Z" near the column;

[0036] S6. Use deep learning methods to obtain the column attribute table location, obtain the location column number and attribute information;

[0037] S7. Obtain the attributes of the column by matching the column number with the column number in the table.

[0038] The data preparation in step S1 specifically includes: 30 residential or basement wall column flat construction drawings;

[0039] The graphic elements in step S2 include visible basic elements that constitute the graphics, which correspond to visible entities on the drawing interface, such as straight lines, arcs, circles, etc. These basic elements constitute meaningful components, such as plates, columns, etc.

[0040] Among them, the space of all columns in step S4 is filtered through the hatch component bbox range to filter out the space columns with interference;

[0041] Among them, in step S5, the column number can be obtained by filtering through regular rules, such as "KZ1";

[0042] Among them, the table detection model trained by Ret i naNet used in step S6.

[0043] The present invention provides a structural professional column attribute recognition method based on traditional image processing combined with deep learning. By applying the traditional image processing method, the problem of accurate identification of components in CAD architectural design drawings is solved, and the excessive dependence of component identification on layers is reduced. By using the deep learning target detection method to identify the column attribute table, the confusion of table style and position can be effectively solved, and the accuracy and speed of obtaining the table position through logical judgment can be improved.

[0044] The above description is only a preferred embodiment of the present invention and 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.

Claims

1. A method for identifying the properties of structural professional columns based on traditional image processing combined with deep learning, characterized in that: The method comprises the following steps: S1. Prepare data first; S2, parsing the drawing to obtain a series of graphic elements; S3, get all the graphic element lines in the column layer, and draw these graphic element lines on a blank base map; S4, using traditional image processing methods to find the space of all columns; S5. Use traditional image processing methods to find text information with "Z" near the column, and filter it through regular rules to obtain the column number; S6. Use deep learning methods to obtain the column attribute table location, obtain the location column number and attribute information; S7. Obtain the attributes of the column by matching the column number with the column number in the table.

2. According to claim 1, a method for identifying structural professional column attributes based on traditional image processing combined with deep learning is characterized by: The data preparation in step S1 specifically includes: 30 residential or basement wall and column flat construction drawings.

3. According to claim 1, a method for identifying structural professional column attributes based on traditional image processing combined with deep learning is characterized by: The graphic elements in step S2 include visible basic elements that constitute the graphics, which correspond to visible entities on the drawing interface.

4. According to claim 1, a method for identifying structural professional column attributes based on traditional image processing combined with deep learning is characterized by: In step S4, the space of all columns is filtered out through the hatch component bbox range to filter out the spatial columns with interference.

5. According to claim 1, a method for identifying structural professional column attributes based on traditional image processing combined with deep learning is characterized by: The table detection model trained by Ret i naNet used in the step S6.

Images