Open shivamvats opened 8 years ago
@shivamvats Lane detection involves the following : data pre-processing => shadow removal => inverse perspective transform => grass removal => obstacle removal => lane detection So do we make a separate class for each of them? Also the functions seperateLanes() and fixBrokenLanes() have been declared within the class but not defined anywhere.
@todivasudha Aren't shadow removal, IPT etc types of data pre-processing? What I am thinking is to have two base classes - pre-processing and lane detection. Rest can inherit from these. The idea is that if later I want to use a different pre-processing technique, I don't have to change any code, but I can inherit from a base class.
Do you have any ideas for base-classes?
minimal
Lane detection and grass removal are logically separate and lane detection need not know about how grass removal is done. Hence, there should be a separate abstract class for data pre-processing, from which grass removal will inherit.