Pigs counting

This is an implemention of our paper "Automated pig counting using deep learning" （https://www.sciencedirect.com/science/article/pii/S0168169918313176）. Our code is based on gramuah's code CCNN. Many thanks to Daniel Oñoro-Rubio and Roberto J. López-Sastre for their codes.

Introduction

Based on this code, it is improved to make it more suitable for pig counts. We have provided the test code, you can test it directly, and all the code will be published after the article is accepted.

Citing Counting CNN

@inproceedings{onoro2016,
    Author = {O\~noro-Rubio, D. and L\'opez-Sastre, R.~J.},
    Title = {Towards perspective-free object counting with deep learning},
    Booktitle = {ECCV},
    Year = {2016}
}

@article{TIAN2019104840,
    author = "Mengxiao Tian and Hao Guo and Hong Chen and Qing Wang and Chengjiang Long and Yuhao Ma",
    title = "Automated pig counting using deep learning",
    journal = "Computers and Electronics in Agriculture",
    year = "2019",

License

The license information of this project is described in the file "LICENSE.txt".

Requirements: software
Requirements: hardware
Basic installation
Demo
How to reproduce the results of the paper
Remarks
Acknowledgements

Requirements: software

Use a Linux distribution. We have developed and tested the code on Ubuntu.
Requirements for Caffe and pycaffe. Follow the Caffe installation instructions.

Note: Caffe must be built with support for Python layers!

# In your Makefile.config, make sure to have this line uncommented
WITH_PYTHON_LAYER := 1

Python packages you need: cython, python-opencv, python-h5py, easydict, pillow (version >= 3.4.2).

Requirements: hardware

This code allows the usage of CPU and GPU, but we strongly recommend the usage of GPU.

For training, we recommend using a GPU with at least 3GB of memory.
For testing, a GPU with 2GB of memory is enough.

Basic installation (sufficient for the demo)

Be sure you have added to your PATH the tools directory of your Caffe installation:
```
export PATH=<your_caffe_root_path>/build/tools:$PATH
```
Be sure you have added your pycaffe compilation into your PYTHONPATH:
```
export PYTHONPATH=<your_caffe_root_path>/python:$PYTHONPATH
```

Demo

We here provide a demo about predicting the number of pigs in the test images of the Pigs dataset.

To run the demo, these are the steps to follow:

Download the our pretrained model.

./our_scale/ccnn_trancos_iter.caffemodel

Finally, to run the demo, simply execute the following command:
```
./tools/demo.sh
```

How to train and test

We provide here the scripts needed to train and test the models with the pigs dataset. These are the steps to follow.

Download a dataset

[Pigs dataset] (https://pan.baidu.com/s/1-iCOU19XewCAKW2TOSiDVw): Extraction code：udei

When the paper is accepted, we can open source data. Put the downloaded train data in counting-pigs/genfiles/features/train, and put the validation data in counting-pigs/genfiles/features/validation.

Download pre-trained models

All our pre-trained models can be downloaded using the corresponding script:

./tools/get_trancos_model.sh

Train model

run the script ./experiments/scripts/trancos_train_test.sh

Test model

run the script ./experiments/scripts/trancos_test_pretrained.sh

Remarks

In order to provide a better distribution, this repository unifies and reimplements in Python some of the original modules. Due to these changes in the libraries used, the results produced by this software might be slightly different from the ones reported in the paper.

wwgc/counting-pigs