Emergency Medical Assistance by Ambulance Drone Using Machine Learning, Light-Weight Cryptography and Variable Image Steganography

A huge amount of people suffer or eventually die every day waiting for the rescue vehicle to reach the spot on time. The primary cause for this delay is traffic jams, blockage in streets, inaccessibility of ambulance and so forth. An emergency ambulance drone or UAV (Unmanned Aerial Vehicle) could be a better feasible solution to this problem travel the aerial path and is unaffected by any terrestrial obstacles, as proposed in this paper. The ambulance drone would be controlled by a human sitting in the hospital control room. This would be the quickest possible medical assistance in case of an emergency. The drone would carry medical aid and using the patient's GPS location, which it receives from the mobile application, can reach the spot without any hindrance. Thereafter the drone can start the audio and video streaming of patient's data to the hospitals, medical centers and necessary instructions can be communicated to the nearby spectators or local doctors for using the right equipment (ECG, Pulse Sensor, Oximeter, etc) for patients diagnosis and detect the face with Haar cascade classifier and recognize the face with LBPH algorithm. In this, a comparative study has been done over face recognition between still and video images, the average accuracy of still and video images is 99.4% and 99.34% respectively. Also, for the situation, when doctors are not present in the hospital, an autonomous system has been designed with the help decision tree that can determine the probable cause of the disease by the patient. The secure information transfer is done via lightweight cryptography and Variable MSB – LSB Image steganography.