Runway holding position signs detection and recognition
is divided into two functional modules - sign detection and
sign recognition. Since runway holding position signs would
only occur in a fraction of the total flight and ground operation
of an aircraft, applying sign recognition using OCR w/o sign
detection would be both computation-inefficient and energyinefficient. Therefore, we would apply runway holding position
signs detection first to locate the sign and then send the
subframe containing only the sign for recognition. If the image
frame does not contain the runway holding position sign, it
would not be sent for sign recognition, thus saving computation
and energy.