The Unmanned Aerial Vehicle team of Engineering Service Learning at UC Merced is working to design and implement a UAV (unmanned aerial vehicle) system that will provide local, small-scale farmers with an efficient method of disease detection and crop management through precision techniques. By designing an autonomous aerial device and a post-flight image analysis program from inexpensive parts and materials, the Unmanned Aerial Vehicle team hopes to reduce the cost to the farmer who wishes to employ precision techniques to detect and stop the spread of disease in perennial crops. The UAV in question will monitor crops and extract meaningful data with minimal input from the consumer, allowing farmers to optimize crop yield and preserve resources in a timely and efficient manner.
The camera’s gimbal has been completed and is ready for testing. It allows the camera to be attached via the 3D printed case to the quadcopter, and face towards the ground. The camera has been set up in such a way as to automatically take pictures at a time interval given by a changeable script.
The image stitching portion of the program, which uses Microsoft ICE, has been improved upon. Major progress has been made on the analyzation portion of the program. The algorithm of this portion is now optimized to use NDVI over RGB. The algorithm is able to distinguish vegetation over non vegetation and distinguish the magnitude “greenness” of vegetation. However the values analyzing color do need to be fine-tuned to distinguish the discoloration of Pierce’s Disease, which are in hues of red, yellow, or orange. An interface has also been created to easily run the program through each stage of stitching and analyzation.