Browse Prior Art Database

Robotic Screw Feeder Using Vision Sensing

IP.com Disclosure Number: IPCOM000049247D
Original Publication Date: 1982-Apr-01
Included in the Prior Art Database: 2005-Feb-09
Document File: 3 page(s) / 54K

Publishing Venue

IBM

Related People

Griffith, JE: AUTHOR [+2]

Abstract

Described herein is a low cost, efficient means for using machine vision to provide a robot system with oriented screws.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 3

Robotic Screw Feeder Using Vision Sensing

Described herein is a low cost, efficient means for using machine vision to provide a robot system with oriented screws.

A standard means of orienting and feeding screws for manual assembly is to use a slotted platform, as shown in Fig. 1. The operator drops many screws on the platform and then jiggles the platform until all of the screws have fallen into slots with the screw heads on top. Subsequently, the operator can insert a tool into the screw slots to pick each one up and use it.

It has been found that the same device (Fig. 2) could be used in conjection with a video recognition system to determine where the screws are on the slotted platform so that a robot can insert a tool into the screw slots to pick each one up and use it. The slot orienter would be placed on a light table. The video camera would see a thresholded pattern of white and black pixels (picture elements).

Fig. 3 shows an example of an actual pattern of screws and the corresponding image seen by the video camera.

The image seen by the video camera may be processed by the computer to calculate the locations of a screw. Although there are several scanning patterns that could be used to calculate the location of a screw, only the most direct will be described here.

Inasmuch as the computer has in its storage an accurate representation of the holder and its slots, it can compare certain features of the image with said representation. Starting with the left end of the top slot (as shown in Fig. 1), the computer scans the slot until it detects a point where no light shows through. If this point does not coincide with the known location of the right end of the slot, it is the location of the left edge of a screw head.

This information, corrected for the distance to the center of the screw head, is then read out of t...