Object Recognition

The aim of this practice is to recognise objects using artificial vision with Vision Builder.

Materials we need:
-1 Computer with Vision Builder
-1 Webcam
-Some objects to recognise (clips, beans...)
-A difuse light

The first thing that we have to do is to put some objects to identify over a difuse light. Then, we will put the webcam over it and we open Vision Builder. Once there, we start the adquisition of images, we convert the image to a greyscale image taking from it the red pannel. Then, we will detect objects and then, we will compare those objects to a database created by us.

This process is similar to the OCR process. We click on Classify Objects and we modify the Threshold to find the objects. Then, we edit the classifier file.

Once in the Classifier File, we select the objects and we teach them to the program.

Finally, we print everything on the screen with a Custom Overlay and we can see a result like this one:

You can download the program from here:

https://drive.google.com/folderview?id=0B20kHfrc3NnpdWhLTzlTdndCVUk&usp=sharing