Скачать с ютуб Pedestrian Controlled Traffic Signals - Final в хорошем качестве

Pedestrian Controlled Traffic Signals - Final

This is the final video for this project - Pedestrian Controlled Traffic Signals. The previous videos in the series were about the lead up with unit testing for the LED traffic signals and Grove Vision AI V2 module and how to prepare and use firmware to deploy in a Xiao ESP32-C3 computer. In this project the Xiao ESP32-C3 get the AI data from the Grove Vision AI V2 module and uses the AI box size number as the count of people present before the Raspberry Pi camera. The Xiao controls the timing of the traffic signals timing and cycle. If more people are present then less time is giving to traffic and less waiting time for pedestrians and more time to cross the road. The pedestrian also get a count down timer and this is displayed on a Grove 4 Digit display. There are two parts for the software. The firmware for the Grove Vision AI V2 has already been packaged by Seeed Studio and is available for immediate deployment to be uploaded as firmware. Alternatively I could have selected a different dataset or created my own and this would result in object recognition with higher levels of confidence and accuracy. I am just using the pretrained model and uploading it to Grove Vision AI V2 using SenseCraft - an vision AI tool created by Seeed Studio. The rest of the software was constructed in the Arduino IDE using C++ with two libraries from Seeed Studio. One is for the Grove Vision AI module and the other is for the Grove 4 Digit display. I am happy with the result and the inference time to detect pedesrian is fast - less than 100 millisecs. I enjoy developing this demo project and hope it becomes useful to others especially for STEM education. My next project will be using the Grove Vision AI V2 module to count cars and this will be used as one of the parameters to control traffic signals.