Project
# | Title | Team Members | TA | Documents | Sponsor |
---|---|---|---|---|---|
61 | Automated Wildlife watcher |
Edwin Lu Kelvin Chen Xu Gao |
Abhisheka Mathur Sekar | design_document1.pdf final_paper2.pdf photo1.jpg photo3.png photo5.jpg presentation1.pptx proposal2.pdf video1.mp4 video |
|
# Title Automated Wildlife watcher Team Members: - Kelvin Chen (kelvin3) - Edwin Lu (jiajun3) - Xu Gao (xugao2) # Problem Despite interests and concern over climate change and human development, there is actually very little data available about both the diversity and distribution of wildlife insects or avian pollinators. This is especially concerning when considering the myriad number of species that are poorly understood. How many are there? How do they live? What do they eat? What can be done to help further their numbers or have the least negative impact. It typically takes a lot of time and effort to survey wildlife populations, a more popular approach is to citizen science. By setting up feeding stations or flowering plants in private residences and documenting visiting species, we can gather a more complete picture of the ecological distribution and possible human impact on the local species. But this too is a limited approach as it depends on observers spending time outside and physically observing and document what they saw, a costly and arguably, ineffective method of data collection. # Solution Our proposed solution is an automated camera system that keeps watch of a specific location, such as a backyard or a patch of flowers, for a prolonged period of time and captures photos or videos of wildlife that enters its view. Because of the proposed size of the area and the smaller relative size of the bird/insect, the camera must be placed on a self-adjustable gimbal that will angle the camera to the bird/insect and so the camera can zoom onto it for a more clear image. This will create a feedback loop of detecting motion, adjusting to the movement, and capturing the movement. # Solution Components ## Subsystem 1: Camera module Camera module with a motion sensing algorithm reacts to dynamic objects (birds, insects, etc.). It has software implemented that is trained to recognize the objects in different directions. When a moving object is detected, the camera module will align and focus on a small area around the moving object and try to follow it using object tracking algorithms like YOLO, Faster R-CNN. ## Subsystem 2: Gimbal stand A gimbal is connected to the camera to stabilize and support it. Once the camera identifies the target object, the motor will turn the camera so that the target will stay within the camera range. ## Subsystem 3: Microcontrolller on a PCB The microcontroller on the customized PCB will be able to receive the data from the camera module and send a signal to the mechanical system. ## Subsystem 4: Power system A power system will be connected to the other subsystems. A voltage converter may be needed to supply the electric energy for the camera module and the gimbal. # Criterion For Success - Camera can detect object entering its field of vision - Gimbal can adjust and follow the object that is moving - The software will zooming the object and capture a photo or video |