Python Image Vision library/framework

3 pointssmithmayowa2 comments

Which image vision library/framework do you think is best for the python programming language for object recognition(i.e recognizing people's body). Thinking of using opencv, is it good enough?.

View on HN

It depends on what you want to do. Opencv is amazing but it doesn't have easy to do high level stuff. It gives you the tools to implement computer vision systems, but you have to put the pieces together.

opencv does not come with models to detect objects and body parts out of the box(except for face detection), so you would have to find tutorials and learn how to do that recognition in opencv.

Building high accuracy detection systems is not trivial, so I would first look into apis from strong companies to see what is available. Start by googling "body part recognition API" and "object recognition API"

Some great detection apis are Google cloud vision, Amazon Rekognition, Clarafai, and there are a few others.

I agree - although I think there are three options:

1. Deep learning - therefore any deep learning framework 2. "Classical AI" - use OpenCV for most of it 3. APIs as you mentioned above

Reply on news.ycombinator.com