Skip to content

A high-fps and pure image processing algorithm for quailty control on mirror production.

License

Notifications You must be signed in to change notification settings

electricalgorithm/QualityConrolMirrorProduction

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Image Processing Method of Quality Control in Mirror Production

Algorithm in detail

Description

This script introduces an image processing algorithm which detects reflection lines on a mirror to find structural faults (such as cambering, deorientiaton, etc.). The algorithm can use its input stream from only one camera, and the camera's positioning is not that important.

The algorithm relies on reflecting uniformed sloped lines into a mirror which will be tested. Due to the camera and environment settings, we have set the default "line detection" amount for calculating average slope as two. For desire of more accurate average angles, please update the minimum line detection amount to an increased number.

  • Tile structure analysis: It allows you to define enclosed areas to find average angles for each-area individually.
  • Debug and log support: Within the debug feature, you can see the chances made in each step of the algorithm both as text, and images.
  • High FPS: The algorithm operates on 30-40FPS on average. Thanks to its easy-to-understand mechanism, the FPS can be increased easily with small adjustments.

Usage

[TODO: Detailed usage] Nam malesuada egestas eros, vehicula varius lorem auctor eu. Nunc dolor nulla, elementum quis sem quis, ultricies elementum nibh. Donec auctor commodo odio lacinia egestas. Duis a lacinia leo. Curabitur sem risus, sollicitudin quis lacus et, varius accumsan urna.

~$ python3 src/detector.py -i path/to/image.png -t 5 2 -lf logs.txt -sd results/dir/path/

Contribution

All contributions are welcome. You can find the guidelines in CONTRIBUTING.md.

License

This project is under the GPL-3.0 license.