THIS PROJECT PROPOSES TO DEVELOP AN ARTIFICIAL INTELLIGENCE (AI) TECHNIQUE CAPABLE OF PROCESSING IMAGES, DETECTING AND TRACKING ALL DROPLETS APPEARING ACROSS THE IMAGE FRAMES, AND MEASURING THE DROPLET SIZE AND MOTION. OUR CENTRAL HYPOTHESIS IS THAT THE INTEGRATION OF DEEP-LEARNING TECHNIQUES INTO THE IMAGE PROCESSING ALGORITHM WILL ENABLE PRECISE AND RELIABLE DETECTION, TRACKING, AND MEASURING OF DROPLET MOTION AND SIZE. IN ADDITION, THE FRAMEWORK WILL PRODUCE CONSISTENT RESULTS UNDER A VARIETY OF UNCERTAIN IMAGERY CONDITIONS. THE RATIONALE IS THAT DEEP LEARNING EXTRACTS MEANING FROM IMAGERY DATA AND HUMAN-LABELED DATA TO TRAIN A LEARNING SCHEME THAT INCREASES THE RELIABILITY AND ACCURACY OF DROPLET TRACKING.WE PLAN TO TEST THIS CENTRAL HYPOTHESIS BY PURSUING THE FOLLOWING THREE SPECIFIC AIMS: 1) DEVELOP A DEEP-LEARNING FRAMEWORK FOR DROPLET DETECTION WITH A FAST PROCESSING RATE, 2) INTEGRATE A FILTERING ALGORITHM INTO THE DEEP-LEARNING FRAMEWORK FOR DROPLET TRACKING, AND3) DESIGN AND IMPLEMENT THE METRICS TO ASSESS THE SUCCESS RATE OF DROPLET DETECTION AND TRACKING.THE PROJECT OUTCOMES WILL PROVIDE AN AI SOLUTION TO A CHALLENGING PRECISION AGRICULTURE PROBLEM, NAMELY MEASURING THE DYNAMIC PROPERTY OF DROPLETS FROM A CROP SPRAYING SYSTEM. THE TOOL WILL LEAD TO THE NEXT GENERATION OF AGRICULTURAL NOZZLES WITH SIGNIFICANTLY IMPROVED PERFORMANCE, WHICH WILL IMPROVE PRODUCER EFFICIENCIES AND MINIMIZE CHEMICAL RUNOFF THAT POLLUTES THE ENVIRONMENT.