Extracting Wildlife in Puma Project Photos Using Java

Overview

Objectives/Goals As part of the Summer Internship Program in Environmental Studies, at the University of California Santa Cruz last summer, I manually tagged a number of images which had wildlife in them. As this was time consuming, I was curious if I could automate this process. My objective is to develop a Java program to extract changes in images from the Puma Project footage to help identify wildlife found in the Santa Cruz Mountains. This program compares the pixels between two successive images in the Puma footage stream to highlight the presence of animals. Methods/Materials The Puma Project footage was obtained using motion sensitive cameras monitoring wildlife. The program to extract the animals (from two successive images) was written in Java, using the Eclipse platform. The user interface was created using JavaFX with Scene Builder 2.0. Four steps were coded to extract changes in the images. 1. Converting image(s) to greyscale. 2. Creating a new image that was the 'difference' of two input (color/greyscale) images. 3. Creating an adjustable filter to mask out the identical portions of the images 4. Applying the mask to the original image to extract the wildlife. Results I was able to extract the animal from the background images using my program. I ran the program on 20 sets of the images from the Puma Project footage. The masked image provided the best results. The mask value that provided the best extracted image varied across images. It was not easy to recognize animals in the 'difference' image. Operating on greyscale versions of the images did not change the results significantly. Conclusions/Discussion My program demonstrates that it is possible to automate the tagging process from two successive images. It met my objective for this project, where my program using Java can extract an animal/object from a background image.

Competition history

• CSEF 2018 Computational Systems & Analysis · Entry S0821

Resources

• CSEF project summary ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Systems Software and Computer Science projects
• CSEF Systems Software and Computer Science projects
• Every CSEF 2018 project