Crowd-Counting-With-LBP

Prerequisites

• Python 3.6
• SciPy library
• SciKit image library
• OpenCV library
• SKlearn library

Quick start

Download the ShanghaiTech Database by looking into the repository and put it in the root folder

https://github.com/desenzhou/ShanghaiTechDataset

Run the extractDescriptorsAndGt.py file to extract lbp descriptor and groundtruth and save them in mat files

python extractDescriptorsAndGt.py --data_root="../ShanghaiTech/ShanghaiTech/part_B" --showImg=1

Run the trainAndTestLBP.py file to train and test different parameters from the ridge regression (alpha and gamma)

python trainAndTestLBP.py --data_root="../ShanghaiTech/ShanghaiTech/part_B"

Run predictionsLBP.py file to train the ridge regression with suitable parameters and save results of the predictions as images

python  predictionsLBP.py --data_root="../ShanghaiTech/ShanghaiTech/part_B" --alpha=0.001 --gamma=0.0036