Intellectual Property India Publishes Patent Application for Ai-Based Engineering Drawing Dimension Identification And Part-Dimension Correlation System

MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202641082841 A) filed by Immanual R; R Kumar; Kavitha Naickenpalayam Shanmugavadivel; Tamilselvan Rathinasamy; Sanjana N; and Alan Sahayaraj on July 06, 2026, for Ai-Based Engineering Drawing Dimension Identification And Part-Dimension Correlation System.

Inventors include R Kumar; Kavitha Naickenpalayam; Tamilselvan Rathinasamy; Immanual R; Sanjana N; and Alan Sahayaraj.

The application for the patent was published on July 10, 2026, under issue no. 28/2026.

Abstract: An AI-based engineering drawing dimension identification and part-dimension correlation system (100) comprises a web-based user interface (102), an image pre-processing module (104), a view separation module (108) comprising an angle of projection classification network (110), a dimension detection module (112) comprising a dimension detection neural network (114), a title block extraction module (116), a GD&T symbol recognition module (118) comprising a vision-language model (120), a correlation and validation engine (122), a recommendation module (124), and a report generation module (126). The view separation module (108) determines the angle of projection of an uploaded engineering drawing (130) and segments it into separated views (132), while the dimension detection module (112) detects dimension lines and numeric values within each view, returning a detected dimension (134) with a confidence score (138). The GD&T symbol recognition module (118) detects and classifies geometric dimensioning and tolerancing symbols (136) in accordance with a referenced standard. The correlation and validation engine (122) correlates the detected dimensions, symbols, and extracted title block metadata with a part type and flags anomalies, while the recommendation module (124) suggests appropriate tolerance parameters. The report generation module (126) compiles a structured inspection report (128). The system achieves dimension detection precision exceeding ninety percent and a processing time of thirty to forty-five seconds per drawing, making automated engineering drawing interpretation accessible across design, quality, and production workflows.

Disclaimer: Curated by HT Syndication.