Visual Pattern Test Inspection of Gauges

Document Type

Project

Lead Author Type

CIS Masters Student

Advisors

Dr. Jerry Scripps, scrippsj@gvsu.edu

Embargo Period

9-15-2016

Abstract

In this project I am taking a visual test program designed to work in a manufacturing environment and add the ability to recognize images with minimal user input. Using a Sobel filtered image I use a series of K-Means and Nearest Neighbor algorithms to determine objects and edges. With this information I created a distance calculation that defines the image based on where objects and edges are, what color, and how many objects. This distance is compared to all known images and the closest image is the matching pattern.

This document is currently not available here.

Share

COinS