Extracting the image of the fingerprint

What is needed is that you have experience in extracting the image of the. Fingerprint with a plug-in of some library that calls the macro function to see the most microscopic of the valleys and ridges that make up the fingerprint. Some cell phones have macro function. The library that is needed must contain a neural network that analyzes when the footprint with the perfect magnification is in perfect focus. To do this, this neural network will analyze the minimum fps of a basic camera and then it is processed to identify that it is a 3D image and convert it into a 2D plane in order to certify that it is a live finger and not an image to reject possible fraud. It will then be converted to grayscale at a 500dpi format for comparison with a stored fingerprint. The library must be extremely fast for this process. That is why the publication is C++ language To cut processing times and 1 or 2 neural networks to process the quality and image capture and the certification that it is a real live image