Event Title

Automated Visual Inspection for Discriminating Between Heraldic Devices and Text

Location

CSU 202

Start Date

23-4-2007 3:00 PM

End Date

23-4-2007 5:00 PM

Student's Major

Computer Information Science

Student's College

Science, Engineering and Technology

Mentor's Name

Rebecca Bates

Mentor's Department

Computer Information Science

Mentor's College

Science, Engineering and Technology

Description

Members of the Society for Creative Anachronism (SCA) distinguish themselves through heraldic devices, similar to those used in the middle ages. Members submit a drawing of their heraldic artwork, such as that found on coats of arms and family crests, which is then compared to all previously submitted drawings before approval. Because these devices need to be distinct from all previous ones, it would be useful to have an online catalog for members to refer to in preparing their devices. When creating the catalog, it is important to select only the device from a scanned form because the application forms contain personal information that should not be placed on the web. In order to meet this need, we have developed a system for selecting the picture of the device from a scanned image of the application form. We will present a comparison between an edge-detection-based approach and an approach based on optical character recognition (OCR). The edge detection approach finds the edges of the device and selects only the portions of the image within a large, connected edge. The OCR approach selects the largest single "character" in the file, which would be an unrecognized character that contains the image. Comparisons of program size, cost, and extraneous pixels included or excluded will be presented.

This document is currently not available here.

Share

COinS
 
Apr 23rd, 3:00 PM Apr 23rd, 5:00 PM

Automated Visual Inspection for Discriminating Between Heraldic Devices and Text

CSU 202

Members of the Society for Creative Anachronism (SCA) distinguish themselves through heraldic devices, similar to those used in the middle ages. Members submit a drawing of their heraldic artwork, such as that found on coats of arms and family crests, which is then compared to all previously submitted drawings before approval. Because these devices need to be distinct from all previous ones, it would be useful to have an online catalog for members to refer to in preparing their devices. When creating the catalog, it is important to select only the device from a scanned form because the application forms contain personal information that should not be placed on the web. In order to meet this need, we have developed a system for selecting the picture of the device from a scanned image of the application form. We will present a comparison between an edge-detection-based approach and an approach based on optical character recognition (OCR). The edge detection approach finds the edges of the device and selects only the portions of the image within a large, connected edge. The OCR approach selects the largest single "character" in the file, which would be an unrecognized character that contains the image. Comparisons of program size, cost, and extraneous pixels included or excluded will be presented.

Recommended Citation

Marte, Ivan A.. "Automated Visual Inspection for Discriminating Between Heraldic Devices and Text." Undergraduate Research Symposium, Mankato, MN, April 23, 2007.
https://cornerstone.lib.mnsu.edu/urs/2007/oral-session-08/6