I want to make a auto screen capture program and the program recognize each questions' height and width and control the capturing size (like green box in example image).
Is the deep learning or AI technology essential?? I really want to make this program but I don't know what language should I study.
No you don't need deep learning. if you environment (where your images are placed, maybe it is full screen, or not) is controlled, you can leverage simple image processing techniques. If you don't have such knowledge, using a pre-trained object detection and object localization libraries (maybe using deep-learning) would be the next choice.
