Differences

This shows you the differences between two versions of the page.

--- emrp2022:start [2023/03/19 06:03] – f osama-haiyl-attallah.attallah
+++ emrp2022:start [2023/03/19 07:07] (current) – osama-haiyl-attallah.attallah
@@ Line 709: / Line 709: @@
 - Brake mode: Immediately stops the drone
-Try to setup a logical and easy-to-configure choice of flight modes, as it will be necessary to remember how you mapped them later during flight. The current set flight mode appears on the main screen of the ground control software.
+Try to make a logical and easy-to-configure choice of flight modes, as it will be necessary to remember how you mapped them later during flight. The current set flight mode appears on the main screen of the ground control software.
 Note while configuring the remote to the Qgroundcontrol: while doing subtrim process and setting up flight mode DO NOT SET THE EMERGENCY STOP MODE ON, it will crash the drone.
@@ Line 746: / Line 746: @@
 ==Setup power and ESC calibration==
-Before setting up the power source, some essentials have to be provided:
+Before setting up the power source, some considerations:
@@ Line 811: / Line 811: @@
 {{youtube>tCXYo-r2UPo?}}
-=====Programming and Control=====
+=====Programming, Control and Drone Applications=====
-Drones can be programmed through a software development kit (SDK) or application programming interface (API) in their packages. Most of the drones that are programmable already come with their own firmware. Specialized functions or applications can then be be developed with a programming language like Python, or the SDK if already provided. Dronekit is an API that can be used to interface the drone by sending MAVlink commands to the drone. Dronekit comes readily equipped with intelligent flight path planning and other autonomous functions All complex tasks must be done from the source of a different framework, such as MAVSDK or ROS (robot operating system), which is a group of modular frameworks, wherein sensors publish data to subscribers. ROS and ROS 2 are different systems with the same concept. MavROS is a direct connection interface that pushes all the info from MAVlink into a ROS format.
+Drones can be programmed through a software development kit (SDK) or application programming interface (API) in their packages. Most of the drones that are programmable already come with their own firmware. Specialized functions or applications can then be be developed with a programming language like Python, or the SDK if already provided.
+Dronekit is an API that can be used to interface the drone by sending MAVlink commands to the drone. Dronekit comes readily equipped with intelligent flight path planning and other autonomous functions All complex tasks must be done from the source of a different framework, such as MAVSDK or ROS (robot operating system), which is a group of modular frameworks, wherein sensors publish data to subscribers. There is also ROS 2, but ROS and ROS 2 are different systems with the same concept. MavROS is a direct connection interface that pushes all the info from MAVlink into a ROS format. Drones have a the vast potential in remote sensing because of aerial photographic and surveillance capabilites, with algorithms from deep learning like classification, localization, object recognition etc.
-For image processing, you can't directly connect the camera to the flight controller, for that you need a companion computer. Offboard mode needs to be enabled when you link to a raspberry pie or jetson.
+However, there are some limitations that can hinder drones. First is the restriction on the weight and area available on the drone surface necessary to compute alogrithms with few hardware resources. If adequate processing power is needed, at most a battery with more cells i.e. bigger voltage and weight are also required. Additionally, for image processing, you cannot directly connect the camera to the flight controller, for that you need a companion computer. In addition, offboard mode needs to be enabled when you link to a raspberry pie or jetson, which limits drone movement and control.
+A prominent example of object classification applications is the identification and distinction between dogs and wolves, this can be useful in, for example farms or ranches where cattle lives, where the operating drone spots an invading wolf and would enact a mission program to chase and scare it away. Another instance is the rescue of lost pets, or identification of drowning or endangered animals, to help rescue them. By programming in Python and applying convolutional neural networks, this can be setup with the following:
+Data/Photos selection: a dataset of different dog and wolf breeds is assembled
+Data augmentation: setting images to a certain resolution, generating tensor images
+Preparing and testing the model: training convolutional models
+Using Data with convolutional neural networks
+Evaluation of results: based on classiication accuracy
+For more details: https://www.kaggle.com/code/mohamedadel7774/cnn-for-dogs-and-wolves-with-accuracy-85#Import-Labraries
+Another capability is the appliation of drones in agriculture and environmental monitoring, where drones equipped with cameras can enhance precision agriculture performance, using RGB sensing and vegetation indices interpretation to produce soil analysis figures, assisst in fertiliser and pesticide spraying and crop health monitoring.
+source: https://www.mdpi.com/journal/remotesensing/special_issues/dronesagri
 =====Conclusion=====
-Going through this tutorial, you should have gained enough knowledge about certain components of drones, the step-by-step assembly of a DJI F550 Flamewheel hexacopter, ground control software, calibration of different parts of the drone, some safety precautions to stick to when working with drones, got to know briefly about the programming of drones and their wide application potential
+Going through this tutorial, you should have gained enough knowledge about certain components of drones, the step-by-step assembly of a DJI F550 Flamewheel hexacopter, ground control software, calibration of different parts of the drone, some safety precautions to stick to when working with drones, got to know briefly about the programming of drones and their promising potential in remote sensing applications such as for domestic animals, wildlife diffentiation, agricultural benefits and environmental monitoring
 =====Supplementary=====
@@ Line 830: / Line 851: @@
 {{ :emrp2022:679iph1pgmm41.jpg?600&nolink }}
 <imgcaption image1>https://engineeringmedia.com/map-of-control</imgcaption>
+ Figure 44: Map of control theory
 === Inverted Pendulum with PID ===