Permalink
Browse files

Typos, added item to test section (red) please check. Marked missing …

…references
  • Loading branch information...
1 parent 47a8f1d commit 7b1af7d8b3d18d17bf53432ac6f41ac01ac174fd Jan Sommer committed Mar 28, 2012
Showing with 10 additions and 9 deletions.
  1. +10 −9 uni-space_payload_pdr.tex
View
19 uni-space_payload_pdr.tex
@@ -166,8 +166,8 @@ \subsection{Functional requirements}
\subsection{Technical requirements}
\begin{itemize}
-\item Operate in open air environment up to 800m over ground.
-\item Operate at 5V unstabilized input voltage at a maximum power consumption of 2.5W.
+\item Operate in open air environment up to 800~m over ground.
+\item Operate at 5~V unstabilized input voltage at a maximum power consumption of 2.5~W.
\item Store at least 1000 medium-resolution images.
\end{itemize}
@@ -199,11 +199,11 @@ \subsubsection*{Board computer}
For controlling the sensors and the camera as well as communicating with the ground station first a standard microcontroller was considered, but as the requirements in \ref{sub:Functional_requirements} state, it is necessary to get a picture around every second. Also the resolution of the pictures should be in the range of some megapixels. Taken this into account, a standard microcontroller would not be able to process these amounts of data fast as well as it normally does not come with a USB-interface necessary for connecting most cameras.
-Instead it was decided to use a BeagleBone embedded computer \cite{BeagleBone:SRM}. It is populated with an ARM Cortex-A8 microprocessor running at 500 -- 700 MHz and provides an USB-port, several I{\texttwosuperior}C and serial interfaces. It is capable of running an embedded Linux operation system and therefore able to support a large varity of devices (plugged to the USB-port) and to provide the possibility for sophisticated onboard calculations from a large set of libraries.
+Instead it was decided to use a BeagleBone embedded computer \cite{BeagleBone:SRM}. It is populated with an ARM Cortex-A8 microprocessor running at 500 -- 700 MHz and provides an USB-port, several I{\texttwosuperior}C and serial interfaces. It is capable of running an embedded Linux operating system and therefore able to support a large varity of devices (plugged to the USB-port) and to provide the possibility for sophisticated onboard calculations from a large set of libraries.
\subsubsection*{Sensors}
-For sensors it is planned to use the LSM303 \cite{LSM303:datasheet} combined magnetometer and accelerometer and the ITG-3200 triple-axis gyroscope \cite{ITG-3200:datasheet} from sparkfun. Both sensors have been used during the CanSat-project in Würzburg, hence the group is familiar with working with this sensors. They communicate via an I{\texttwosuperior}C interface with the main board. For receiving GPS information a LS20031 5~Hz GPS receiver \cite{LS20031:datasheet} can be used. It is connected via a serial line interface with the main board.
+For sensors it is planned to use the LSM303 \cite{LSM303:datasheet} combined magnetometer and accelerometer and the ITG-3200 triple-axis gyroscope \cite{ITG-3200:datasheet} both from sparkfun. Both sensors have been used during the CanSat-project in Würzburg, hence the group is familiar with working with these sensors. They communicate via an I{\texttwosuperior}C interface with the main board. For receiving GPS information a LS20031 5~Hz GPS receiver \cite{LS20031:datasheet} can be used. It is connected via a serial line interface with the main board.
\subsubsection*{Transmitter/Receiver}
@@ -212,7 +212,7 @@ \subsubsection*{Transmitter/Receiver}
\subsubsection*{Camera}
-As high quality embedded industrial cameras are very high priced it is planned to connect a consumer webcam with a resolution of several megapixels to the USB-port of the main board. It is intended to buy a camera which is supported by the Linux operating system running on the main board. To save weight, the case of the webcam will be stripped as much possible leaving only the bare camera and electronics. An example of a compatible camera would be the Logitech HD Webcam C525\footnote{\url{http://www.logitech.com/de-de/webcam-communications/webcams/devices/7794}}.
+As high quality embedded industrial cameras are very high priced, it is planned to connect a consumer webcam with a resolution of several megapixels to the USB-port of the main board. It is intended to buy a camera which is supported by the Linux operating system running on the main board. To save weight, the case of the webcam will be stripped as much as possible leaving only the bare camera and electronics. An example of a compatible camera would be the Logitech HD Webcam C525\footnote{\url{http://www.logitech.com/de-de/webcam-communications/webcams/devices/7794}}.
\FloatBarrier
@@ -227,7 +227,7 @@ \subsection{Attitude Determination System}
\end{figure}
In order to produce high-accuracy attitude estimates and compensate for disadvantages of certain sensor types such as drift and noise, we chose to use a variety of sensors and fuse their information to a combined information.
-The facilitated sensors will be (\ref{fig:ADS_overview}):
+The facilitated sensors will be (see figure \ref{fig:ADS_overview}):
\begin{itemize}
\item GPS receiver: Provides absolute position values, but has much high-frequency noise
\item Gyroscope: Provides accurate relative pointing direction, but has drift.
@@ -246,7 +246,7 @@ \subsection{Software environment}
Since a Linux operating system is run as a basic software layer, it is planned to deploy the Robot Operating System (ROS) as a middle-ware. This system is a modern framework
for mobile robotic applications. Its main advantage is the interoperability of modules and inter-use of underlying
layers. This increases the flexibility of future research and overall system robustness. Since ROS is an open-source system along with its packages it is used for knowledge exchange and achievement demonstration in terms of modern algorithms. These algorithms include Simultaneous Localization
-And Mapping (SLAM), Navigation and position control, Image processing and many more.
+And Mapping (SLAM), Navigation and position control, Image processing and many more. This way the IPS could also easily be mounted to different carrier systems and, if desired, be responsible for flight control.
\subsection{Image processing}
The main goal is to take numerous aerial images and combine them into one usable map. This system should then be able
@@ -260,12 +260,12 @@ \subsubsection*{Image matching}
\item Step\\
It is needed to extract common and significant points from all the images in question. Since the images
will be taken in sequence it can be assumed that there will be a high correlation in the neighboring pictures.
-Having this assumption an algorithm proposed in [bib-ref] can be implemented for finding
+Having this assumption an algorithm proposed in {\color{red} [bib-ref]} can be implemented for finding
Points of Interest (POI) based on histogram leveling.
\item Step \\
Having a reliable set of POI another technique to match the points together can be used. This part is
computationally intensive. In order to be able to use it in real-time it is recommended to use the Iterative Closest Point
-algorithm already implemented and optimized in the Point Cloud Library[bib-ref].
+algorithm already implemented and optimized in the Point Cloud Library{\color{red} [bib-ref]}.
\item Step\\
By knowing the difference between two set of points in two consecutive images it is possible to calculate
a very precise transformation matrix. The transformation will be determined by how many points
@@ -282,6 +282,7 @@ \subsection{Design Models and Verification Methods}
\begin{itemize}
\item A Development Model (DM) will be built, using a breadboard to connect the electronic components together. System stability will be monitored by sequence of tests. Modifications of the software will be done if needed.
\item If time allows, a Flight Model (FM) will be build, using a custom designed PCB schematic layout. This design would minimize circuit mass, size and internal losses.
+\item {\color{red} Image processing and stitching algorithms can be tested with a set of example pictures on the ground before using pictures taken by the payload. The performance of example and taken pictures can then be compared.}
\end{itemize}

0 comments on commit 7b1af7d

Please sign in to comment.