Poster Session #1: UC Ballroom
Using Image Detection to Locate Resin Ducts
Presentation Type
Poster
Faculty Mentor’s Full Name
John Bardsley
Faculty Mentor’s Department
Mathematical Sciences
Abstract / Artist's Statement
Computers are an indispensible tool in the lives of many citizens, but that powerful technology can also be harnessed to aid in research. This project focuses on creating a code that is capable to detecting, locating and measuring resin ducts. Resin ducts are the remnants of veins that once transported resin through the bark. They appear as small bright circles that lie in the tree rings and bark of trees. The number of resin ducts and their respective size in a tree can be indicative of the tree's health and ability to protect itself from Bark Beetles, a pest that threatens the health of trees.
Before completion of this project the time needed to locate all the resin ducts and measure their areas in a single slab of a large tree was close to three hours. By using the powerful mathematical processing program, Matlab, a code was created that was capable of searching through an uploaded picture for resin ducts, marking their location within the picture then outputting their respective areas in a fraction of a second. However, it was also designed with the capacity for a human researcher to oversee its function to make critical distinctions between ambiguous areas of the picture. Early versions of the code were tested to be approximately 87%, but in the interest of accuracy sliders were added in order for the researcher to control the sensitivity of the program on a case-by-case basis based upon the quality of the picture. This code can be used for a variety of different studies of resin ducts in order to determine many characteristics of trees and their surrounding environments.
Using Image Detection to Locate Resin Ducts
UC Ballroom
Computers are an indispensible tool in the lives of many citizens, but that powerful technology can also be harnessed to aid in research. This project focuses on creating a code that is capable to detecting, locating and measuring resin ducts. Resin ducts are the remnants of veins that once transported resin through the bark. They appear as small bright circles that lie in the tree rings and bark of trees. The number of resin ducts and their respective size in a tree can be indicative of the tree's health and ability to protect itself from Bark Beetles, a pest that threatens the health of trees.
Before completion of this project the time needed to locate all the resin ducts and measure their areas in a single slab of a large tree was close to three hours. By using the powerful mathematical processing program, Matlab, a code was created that was capable of searching through an uploaded picture for resin ducts, marking their location within the picture then outputting their respective areas in a fraction of a second. However, it was also designed with the capacity for a human researcher to oversee its function to make critical distinctions between ambiguous areas of the picture. Early versions of the code were tested to be approximately 87%, but in the interest of accuracy sliders were added in order for the researcher to control the sensitivity of the program on a case-by-case basis based upon the quality of the picture. This code can be used for a variety of different studies of resin ducts in order to determine many characteristics of trees and their surrounding environments.