Getting Started

Table of Contents

• Getting Started
  • Install OD
  • Get Data
  • Run your first examples

Getting Started

Install OD

First install OD as explained in Installation Instructions. Basically, after installing dependency you will be downloading and building OpenDetection with:

git clone https://github.com/krips89/opendetection.git

cd opendetection 
mkdir build; cd build
cmake ..
make -j8

Get Data

After compiling OD, you will need some data to train or some existing pre-trained data to run any of the examples in the examples directory. To help get started we have provided our own data in the repo opendetection_data. Download our data from https://github.com/krips89/opendetection_data using

git clone https://github.com/krips89/opendetection_data.git

Make <path_to_data>/trained_data as your trained_data location in all the examples, as pre-trained data of some Detectors are already provided with proper directory structure.

Run your first examples

Go to the build directory of OD (<path_to_source>/build). Run a face detector from your webcam using:

./examples/objectdetector/od_cascade_cam <path_to_data>/trained_data/

Or get some images containing people (suppose in <path_to_images>/*.JPG) and run the people detector using:

./examples/objectdetector/od_image_hog_files <path_to_data>/trained_data/ "<path_to_images>/*.JPG"

Depending on the image, you will see something like the following.