======= #This semantic mapping code is my graduation project
- Models files are not here, you can choose to modify this file:
//segnet.cpp
std::string model_file = "../models/segnet_model_driving_webdemo.prototxt";
std::string trained_file = "../models/segnet_weights_driving_webdemo.caffemodel";
std::string label_file = "../models/semantic12.txt";
- The semantic segmentation methon is based on Alex Kendall's work
- Caffe Segnet is a modified version of Caffe which supports the SegNet architecture
- The mapping architecture is based on GaoXiang's work and ORB-SLAM.
If you would like to try out this code , you should satisfy these requirements first.
#OpenCV I choose to use OpenCV to manipulate images and features. Dowload and install instructions can be found at: http://opencv.org. Required at OpenCV 2.4.x.Note that my code is not compatible with OpenCV3.0