High-altitude parabolic detection system based on machine vision

The high-altitude parabolic intelligent detection and early warning system based on machine vision is a high-precision, low-cost, practical, and easy-to-assemble new system designed by combining existing technical means. Combined with traditional video surveillance, the system performs morphological processing on the initial image through grayscale and Gaussian filtering methods, and then transmits data quickly through the 5G communication module. The receiver uses a convolutional neural network algorithm to identify the type of objects falling from high altitudes. When entering the threat mode, while identifying the falling object, through the mathematical model established by analyzing its possible force mode, the Kalman filtering method is used to jointly predict the data received by the dual cameras, and the falling object is predicted. Point coordinates and transfer them. The sound and light early warning module that receives the coordinates quickly processes the data, uses eye-catching laser lights to mark the landing point before the falling object hits the ground, and supplements it with a sound alarm to remind pedestrians, and uses a blocking net to intercept to avoid pedestrians from being hurt. This paper focuses on a high-altitude parabolic detection system based on deep learning and intelligent prediction.