Image Processing is extremely useful for automating tasks a human would otherwise have to perform. This ImageJ plugin automates the task of partitioning x-ray images of shape charge jets. A shape charge jet is a stretching rod of explosively formed metal that eventually particulates. The most famous weapon that makes use of a shape charge jet is the RPG-7. In order to characterize the performance of a jet, x-rays of the jet from different times are measured. In this way, one can compare jets in an empirical way. The plugin works by first partitioning the image into separate flashes by analyzing an image projection after a Sobel edge detection filter is applied. Then each flash is broken into pieces by partitioning the particles in each flash using the same techniques. After the image is broken into flashes in the y direction, then each flash can be broken into particle segments in the x direction. Finally, each particle segment can be further bound in the y direction. This results in a hierarchy of particle bonding boxes organized by position in x and flash. The plugin uses edge detection and image projections to accomplish this task. This plugin was developed as a preprocessing step for further analysis software that generates jet characteristics data from the x-ray images, given that the particle positions are known. Previously, this work was done manually by measuring each particle on the x-ray film, or by manually selecting each particle across all the images. With this plugin, the whole process has been made much easier.
partitioning, edge detection, projection, particle, jet x-ray
I am currently working towards my Master’s degree in Computer Science at the University of Alabama At Birmingham in Birmingham, AL. I am currently interested in the topics of computer graphics, shape modeling, and image processing. My undergraduate honors thesis was on a method to render algebraic curves. I am currently working with shape modeling in conjunction with 3D printing.
Presenting author: Patrick Butler
co-authors: Kenneth R. Sloan, email@example.com, UAB