Extracting T features from an Image

i need to extract T feature points from an image , T features would be comprise of the area of two pair of eyes and nose only , can somebody help me out how to go through to extract these features.

Answer 1

[Anand]

Its not clear to me what exactly you want - feature points, an area or something else.

But this should get you started nevertheless:

Use the vision.CascadeObjectDetector object to get a bounding box around the eyes and the nose. Use one of the "EyePairBig" or "EyePairSmall" classification models for the eyes. For the nose, use the "Nose" classification model.

[Ali Zeeshan]

Thank you , i got your point but now i need to combine these boxes to form a T pattern and extract feature points and store in a metrics . can you help me out how to do that ?

[Anand]

What do you mean by 'extract feature points and store in a metrics'?

[Ali Zeeshan]

I just need to extract features from that T pattern and need to do some analysis on these features so for that i need to store pixel values or anything better than that data, if you can help me out ?

[Ali Zeeshan]

@anand ? can you help me out ?