Project

During my third year of study, I interned at MimoMax Wireless, a Christchurch-based company specializing in the design and manufacturing of Multiple Input Multiple Output Radios. My assignment was to design and develop an automated testing software, simplifying the laborious process of testing new radios. Project Requirements: User-Friendly GUI: Intuitive interface for ease of use. Scalable Architecture: Ability to accommodate new developments seamlessly. VISA Interface Compatibility: Integration with VISA interface for enhanced flexibility....

Read the Final Published Paper The Problem Prolonged parking overstays result in revenue loss for the parking company. The current manual method of detecting parking occupancy is inefficient and requires parking officers. Solution Implement a high-mounted camera system with computer vision algorithms and deep learning to validate parking spot occupancy. Key Steps: Grey Scaling and Gaussian Blur Background Subtraction Thresholding Dilation Non-Zero Pixels YOLO Neural Network Results Methods Accuracy Explanation Background Subtraction Only 50% Detected every small movement Background Subtraction + Thresholding 72% Filtered out small movements Background Subtraction + Gaussian Blur + Thresholding + Dilation 84% Reduced noise from the image; still detected humans and other non-vehicle related movements Addition of YOLO Neural Network 91% Eliminated non-vehicle related movement The final achieved accuracy was 91%, surpassing previous research on this application 1....