Drawing text reading method and system based on cluster analysis
A technology of cluster analysis and drawing, applied in the direction of instruments, calculations, characters and pattern recognition, etc., can solve problems such as translation inconvenience, achieve the effect of improving accuracy and reducing clustering errors
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0037] like figure 1 As shown, the drawing text reading method based on cluster analysis of the present invention includes the following steps: S1: classify the text boxes on the drawing according to their angles; S2: extract the coordinate feature values of the text boxes of the same angle type; S3 : Carry out cluster analysis on the text boxes of the same angle type, so that the text boxes with similar coordinate feature values are clustered into the same class, and sort the text boxes according to the clustering results; S4: sort the text boxes according to the text The angle type of the box for text output.
[0038] When this embodiment is implemented, the text boxes on the drawings are first classified according to their angles, and the text boxes are divided into multiple different angle types, such as 0°, 90°, 180° and 270° that often appear in cad drawings ; Then extract the coordinate feature value of the text box of the same angle type, this coordinate feature v...
Embodiment 2
[0040] In this embodiment, on the basis of Embodiment 1, the coordinate feature value adopts the coordinate value of the upper left corner, the lower left corner, the upper right corner, the lower right corner or the coordinate value of the center point of the text box.
[0041] When this embodiment is implemented, the coordinate feature value adopts the coordinate value of the upper left corner of the text box, the coordinate value of the lower left corner, the coordinate value of the upper right corner, the coordinate value of the lower right corner or the coordinate value of the center point, because the coordinate feature value is to identify each The coordinate value of the unique position of the text box, the above five coordinate values can all express the unique position of the text box, which effectively improves the accuracy of the clustering of the present invention.
Embodiment 3
[0043] In this embodiment, on the basis of Embodiment 1, the clustering adopts the optics algorithm; the optics algorithm determines the relative distance between the text boxes by reading the coordinate feature values of the text boxes in the ordered text box group; The criterion for the similarity of coordinate feature values is that the relative distance is less than or equal to the threshold.
[0044] During the implementation of this embodiment, since the text boxes on the drawings are regular but irregular, the inventor found through creative labor that it is difficult to accurately determine the clustering parameters due to the irregular text boxes, and when using the optics algorithm, the clustering When the parameters change reasonably, the clustering results do not change much. And in the optics algorithm that the present invention applies, determine the relative distance between the text boxes by reading the coordinate feature value of the text box in the ordere...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


