Algorithms for recognition of the main engineering drawing entities

This paper describes a procedure for data extraction and interpretation of section representations in vectorized line drawings. Vectorized drawings are considered as graph representation. The hatch areas are analysed and extracted in regard of their geometric and topologic characteristics with the help of context analysis. With the semantic interpretation, the refered view representations of half-and local sections, sections in same view direction are completed in the viewpoint of orthogonal projection and are then available for 3D geometric reconstruction with known algorithms.

show abstract

“…We call the both as a hatch area A(P,E h ). 5 ), e 6 (v 5 ,v 6 ) and e 7 (v 6 ,v 2 ), only e 6 is hatch line.…”

Section: Extraction Of Section Representationsmentioning

confidence: 97%

Stepwise segmentation and interpretation of section representations in vectorized drawings

Grabowski¹,

Liu²,

Michelis³

1999

Proceedings of the Fifth International Conference on Document Analysis and Recognition. ICDAR '99 (Cat. No.PR00318)

View full text Add to dashboard Cite

show abstract

“…An algorithm to extract arrowheads based only on their shape information was proposed by Ablameyko et al 1,5 A raster-scan approach is used. A line width histogram is computed and local maxima corresponding to thin and thick lines are identified.…”

Section: Arrowhead Extractionmentioning

confidence: 99%

“…1,11 In the case of crosshatched areas, such important and difficult operations like line aligning, changing of positions and number of nodes should be made. Figure 6 shows a section of vectorized ED before and after pruning by the algorithm developed by Ablameyko et al 5,11 One of the first approaches to analyze objects containing regular set of lines (texture lines) was proposed by Kasvand. 43 He studied a general case where no a priori information is available on the texture elements or on the shapes and number of lines.…”

Section: Recognition Of Crosshatched Areasmentioning

confidence: 99%

Recognition of Engineering Drawing Entities: Review of Approaches

Ablameyko

Uchida

2007

Int. J. Image Grap.

View full text Add to dashboard Cite

Recognition of engineering drawing entities is one of the most difficult stages in engineering drawing interpretation and many attempts to recognize various types of ED entities have been made. In this paper, we review algorithms for the recognition of ED entities, especially dimensions and crosshatching areas. For the recognition of dimensions, we analyze how dimension texts can be separated from graphics and how arrowheads of dimension lines are recognized. We also analyze the recent systems of ED interpretation. Finally, future tasks are discussed. Int. J. Image Grap. 2007.07:709-733. Downloaded from www.worldscientific.com by UNIVERSITY OF CALGARY on 02/05/15. For personal use only. Recognition of Engineering Drawing Entities 711arrowheads are recognized. We review complete or near-complete ED interpretation systems. During the past 20 years, many ED interpretation systems have been developed for automatic or semi-automatic conversion of paper-based EDs to 3D CAD model. Some review material can be found in Refs. 55, 66, 70 and 75. We shall not touch in this paper a character (and other specific symbol) recognition task. There are many papers published for recognition of characters and symbols and we can refer to only several review papers published recently. 12,20,25,49 For editing and 3D reconstruction approaches, which are also omitted here, we can refer to Refs. 9, 10, 13 and 32. ED Entities What are the ED entities?Vectorized ED can be represented in two ways: by graphic primitives or by ED entities. The first representation is rather simple and not always useful. Straight lines, arcs, curves are the most commonly used graphic primitives. They can be grouped for forming more complex primitives, such as chained lines, text blocks, and arrows. Those complex primitives may be further combined into higher-level elements, that is, ED entities such as • contour lines (often represented by solid lines), • symmetry axes (dash-dotted lines), • hidden contour lines (dashed lines), • matter areas (represented by crosshatching), • dimensions (thin lines with arrows, witness lines and so on),• annotation text, and • circles including concentric circles.Examples of some ED entities are shown in Fig. 2.ED entities can be roughly separated into simple and complex ones. Simple ED entities are used for the description for simple ED structures, e.g. a symmetry axis, a crosshatching line, the border of matter area, etc. and can be represented as rather simple combinations of graphic primitives (or their pieces). In Fig. 2. Examples of ED entities; (a) dimension, (b) symmetry axis, (c) crosshatching area, (d) border of crosshatching area, and (e) a set of concentric circles. Among them, (a), (c), and (e) are complex entities and (b) and (d) are simple entities. Int. J. Image Grap. 2007.07:709-733. Downloaded from www.worldscientific.com by UNIVERSITY OF CALGARY on 02/05/15. For personal use only.

show abstract

“…This approach differs from well known two or more stage systems, which first identify elements such as lines, arcs or points and perform image interpretation in a second step (see e.g. [6,7]). Such an approach, however, is difficult to follow whenever it is not possible to perform a segmentation of the drawing into the primitive elementary objects as mentioned above.…”

Section: Introductionmentioning

confidence: 99%

Searching an engineering drawing database for user-specified shapes

Rigoll

1999

Proceedings of the Fifth International Conference on Document Analysis and Recognition. ICDAR '99 (Cat. No.PR00318)

View full text Add to dashboard Cite

In this paper we present a novel approach to the retrieval of engineering drawings based on the use of stochastic models. Engineering drawing databases can be searched intuitively by presenting sketches or shapes which represent details such as e.g. screws or holes in the drawings of mechanical parts. The query is represented by a so called pseudo 2-D Hidden Markov Model (P2DHMM) which is surrounded by filler states. These filler states generate the remaining part of the engineering drawing apart from the query shape or sketch itself. Thus, our approach aims to retrieve those images containing certain details and also locates these details in the retrieved images, even in cases where the query shape is embedded in e.g. hatching or is connected to other parts in the drawing. The proposed technique achieves a good performance which is demonstrated by a number of query and retrieval examples in this paper.

show abstract

Algorithms for recognition of the main engineering drawing entities

Cited by 6 publications

References 9 publications

Stepwise segmentation and interpretation of section representations in vectorized drawings

Stepwise segmentation and interpretation of section representations in vectorized drawings

Recognition of Engineering Drawing Entities: Review of Approaches

Searching an engineering drawing database for user-specified shapes

Contact Info

Product

Resources

About