ROMJIST Volume 24, No. 3, 2021, pp. 321-333
Nichita-Maria ȘERBAN, Radu HOBINCU Computer Vision Algorithm for Detecting Resistor Color Codes
ABSTRACT: Identifying resistor color codes for human operators is a relatively simple task, given enough experience so that the colors and positions are memorized. There are, however, difficulties when the color shades are confusing, as well as when the sequence can be read in both directions, not to mention the increased complexity when the resistor has five or six bands, in which case several of them have different meaning and values. This paper presents a computer vision image processing pipeline that attempts to preprocess the image, detect, segment and rotate the resistors, detect and segment the color bands and ultimately identify the nominal value of the resistor, tolerance and temperature coefficient. The results show that the detection is done accurately, if the light conditions are proper.KEYWORDS: computer vision, opencv, color codes, automationRead full text (pdf)
