Optical recognition of electrical circuit drawings

Abstract

This paper presents an optical recognition system for electrical circuits drawing. The system uses computer vision techniques to analyze the circuit's drawing and produces a circuit description program suitable for analysis by circuit analysis programs such as SPICE. The algorithm proposed uses a variety of machine vision techniques to locate the circuit nodes and isolate the circuit components. Moments' invariant descriptors are then used to for the recognition of both elements and values.

Recognition of electrical symbols in document images using morphology and geometric analysis

“…1 • IND An inductor closely resembles RES apropos its periodic strokes, which are, distinguishingly, constituted by spiral lines.…”

“…Some of the approaches are off-line [1,2,6,15] and some are online [9,11,15]. Out of all these existing approaches, however, quite a few have focused on recognition of symbols in an electrical drawing present in the input image of a digitized document after segmenting out the text from the graphics.…”

“…It is assumed that all the symbols must have same number of features invariant to their orientation, size, and location in the entire drawing. The algorithm in [1] is designed only for a DC and not for an AC circuit, and uses moment-invariant descriptors. The online scheme in [15] uses the spatial arrangement identified from the strokes while drawing.…”