Due to the coronavirus outbreak, I can no longer go on site for my Senior Project and have been working from home.
This week I focused on refining the method by which we calculated zebrafish behaviors. Dr. Kanwal and I talked through the entire methodology from tracking the fish to detecting distinct motor patterns within this tracking. As we went through the process, step-by-step, we talked about places of improvement. By the end of the call, I had a list of components to improve which I spent the week working on.
First, I looked into tracking multiple fish. The tracking software we have been using to study the zebrafish has released an update allowing us to track multiple fish. I spent some time researching this new feature and worked on implementing it.
Next, I looked into time intervals. So far, during our analysis of the fish, we have been treating each motor pattern as 500ms long. However, we wanted to see what would happen if we changed that interval and tested time intervals of 300ms, 1000ms, and 2000ms.
Another adjustment I looked into was applying filters to our videos of swimming zebrafish. We noticed that during our tracking of the zebrafish, the quality of the video and differing sizes of zebrafish had an impact on the accuracy of tracking. To solve this problem, I have been looking into filters to create a sharper contrast between the fish and the background for easier tracking.