Fszta / Real-time-object-detection

OpenCv dnn + Caffe model

Real-time-object-detection

Real-time object detection based on caffe pre-trained model

Dependencies

• OpenCv
• imutils

pip install -r requirements.txt

Docker

docker build -t object_detection .

Run the project with pre-trained model :

bash RunDocker.sh

Result

• Achieve 19 FPS with 640x480 resolution

Reference

The model is a caffe implementation of MobileNet-SSD detection network trained by chuanqi305