# How can I build a real-time shuttlecock detection system?

Source: https://zilliz.com/ai-faq/how-can-i-build-a-realtime-shuttlecock-detection-system
Author: zilliz

To build a real-time shuttlecock detection system, you can use computer vision with deep learning. First, collect and annotate a dataset of shuttlecock images in different positions and lighting conditions. 

 Train a convolutional neural network (CNN) or use a pre-trained model like YOLO or SSD to detect and track the shuttlecock. These models can localize and classify the shuttlecock in real-time. Use OpenCV to preprocess the video feed and integrate it with the trained model. 

 Optimize the system for speed and accuracy by fine-tuning the model and employing hardware acceleration, such as GPUs or edge devices like NVIDIA Jetson, for real-time performance.