Mastering OpenCV 3 - Second Edition

Training and visualization

An example program for training the patch models from the annotation data can be found in train_patch_model.cpp. With the command-line argument argv[1] containing the path to the annotation data, training begins by loading the data into memory and removing incomplete samples:

    ft_data data = load_ft<ft_data>(argv[1]); 
    data.rm_incomplete_samples();

The simplest choice for the reference shape in the patch_models class is the average shape of the training set, scaled to a desired size. Assuming that a shape model has previously been trained for this dataset, the reference shape is computed by first loading the shape model stored in argv[2] as follows:

    shape_model smodel = load_ft<shape_model>(argv[2]);

This is followed ...