Work on haar

[jgvictores]

In private repo issue https://github.com/roboticslab-uc3m/tiago-main/issues/9 there were attempts to train haar:

Here you have the positive and negative samples, the .vec file for _opencvhaarcascade and the command lines I used for for the different generation/training steps (only as reference, as the paths are invalid)

The objective of this issue is to track down what went wrong, and develop documentation of what made it work if it turns out to be feasible via haar.

[davidvelascogarcia]

Note: our haar xml files are currently stored at https://github.com/roboticslab-uc3m/vision/tree/develop/share/haarDetection/conf

[davidvelascogarcia]

WIP at https://github.com/roboticslab-uc3m/vision/tree/issue-77-haar/share/haarDetection/conf

[jgvictores]

Entiendo que esto es un duplicado de #77. @davidvelascogarcia ¿Ok cerrar issue y eliminar rama, seguro que no se nos pierde ningún fichero?

[davidvelascogarcia]

Sí, #77 es un duplicado de la branch tiago-vision del pull request. Se pueden borrar sin problema y cerrar la branch issue-77 e issue #77. Todos los cambios que se realizaron en issue-77, han sido pasados a tiago-vision. Acabo de revisarlo otra vez por si acaso, solo faltaba añadir una referencia al README que ya ha sido añadida. e6096a31aeaecbd3e2287ba5e910717c6072925f

[jgvictores]

Cerrando!