Title: A novel mobile application for circuit component identification and recognition through machine learning and image processing techniques

Authors: Shriram K. Vasudevan; Karthik Venkatachalam; R.M.D. Sundaram

Addresses: Department of Computer Science and Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham (University), Coimbatore, Tamilnadu, 641112, India ' Department of Computer Science and Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham (University), Coimbatore, Tamilnadu, 641112, India ' Software Architect, WIPRO Technologies, Bangalore, 560100, India

Abstract: Building an electronic product is complex and needs many steps, guidelines to be followed. From initial phase, till testing the product complexities are numerous. Most of the time in building a product is spent towards simulating the design to check if the proposed design is successful. People with expertise in using simulators only are deployed for this purpose. Although simulation and coding skills are secondary which may not contribute to the core value in the product, it is forced that everyone should learn them. To avoid this, we have come up with an innovative solution through an application in mobile phone that involves machine learning and image processing. Our proposed application enables the user to draw the circuit which he wishes to, once drawn with freehand, the circuit would be recognised and segregation happens where wires and components are split appropriately. Then, each component will be identified through the machine learning algorithms and netlist will be generated, all in seconds with high accuracy.

Keywords: netlist; circuit recognition; machine learning; image processing; de-noising; thinning; connected component analysis.

DOI: 10.1504/IJISTA.2017.088049

International Journal of Intelligent Systems Technologies and Applications, 2017 Vol.16 No.4, pp.342 - 358

Received: 25 Sep 2016
Accepted: 22 Jan 2017

Published online: 20 Nov 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article