Question

Introduce current technological challenges with respect to the engineering design of vision systems for drones and autonomous vehicles. Also, present the underlying physical and engineering principle of autonomous vision systems. How can the existing drone detector technology be improved further? (For instance, you may relate to class material with emphasis on deep learning, neural network, cognition vision features, neuromorphic vision, bioinspired vision and other.)

Answer 1

From the last two decades. drones have emerged itself into multitude technologies, starting from Filming a shot to delivering products to the customers, search and rescue, inspection, security, surveillance, science and engineering, R & D, aerial photography, aerial imaging and mapping, surveying, But still there are a lot of areas where this actually fails to meet the need of the customer , and companies are still working for the development of the quality, efficiency and the size of the drone,,

The rules and regulation for flying of drones actually differ to the area around the world, But still, I need to add the rules for piloting a drone, There should be a license for flying of a drone, which can be got by passing an examination in some countries like Dubai. There are dronprohibiteded zones, where we are not permitted to fly a drones, as it can be used to Human Damage and the Privacy Thefting of personal or country can be revealed by this,

Vision research has uncovered computational principles that generalize across species and brain area. However, these biological mechanisms are not frequently implemented in computer vision algorithms.But the development can gain a good productivity of the autonomous equipments, The brain mimics, the Facial recognisition and the feedback Api can bring a great change with the use of deep learning which are actually the advanced level of the machine learnig,

Some of the limitations of the drone technology which are common in almost any fields are:-

1. The drones battery backup is very less as it carries 4 motors and camera setup with it, the draining of the battery is much higher, which needed to be charged every time we need to use it,

2. The Height of the drone is also the height of the Birds, such as eagles which attacks the drones,

3. Better batteries/better aerodynamics are needed to help drones fly through the entire night without needing recharging, as if this can be done, a person can double their productivity as the drone works while they sleep.

4. drone collisions with aeroplanes is also a major problem, however, keeping track of all the UAVs and their flight routes seems to be the primary concern. Another problem with having UAVs enter airspace is the risk of property damage.

Neuromorphic vision:-

(1) task-specific spatial variant layouts,

(2) continuous time and steady-state phototransduction,

(3) dynamic range compression using logarithmic and adaptive photoreceptors,

(4) early visual processing capabilities for a focal plane edge, motion and orientation detection

BIo inspired vision:- Technology approach of the neuromorphic visual information processing offers an alternative which mimics the robustness and flexibility of the primary visual cortex, based on Hubel and Wiesel’s experimentation and Hodgkin-Huxley formalism. This will help the autonomous and self driving drones, The idea is applied to the detection or recognition of human objects with the head and torso shown, though it can mimic the human or animals’ visual intelligence of recognising any particular objects under the extreme visual environment.

Deep learning are the algorithms which can be used for the autonomous technology which are actually inspired by the brain functioning and the way of solving a problem how a human does, The performance and the output of the Deep learning will definitely help the autonomous vehicles to detect the obstacle and accordingly react without damaging itself and others,

Cognitive Vision:- This technology is actually based on the Recognition and focused on the object, facial, and emotion detection. This will help the autonomous vehicle to determine the size and the Facial recognition of the human, so that the security of the services may be increased.

FEEDBACK API:- These have to be developed in the autonomous vehicles, as it can help the security and the output of the vehicles or drones and accordingly do the job to be done without any damag