pavel/kick-it
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

continued reading

Browse Source
This commit is contained in:
jonas
2018-08-08 23:51:24 +02:00
parent c0c85bf017
commit e8a7040d54
4 changed files with 29 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
documentation/jonas.tex
View File

@@ -45,7 +45,7 @@ camera frames.
In the \textbf{Head Adjustment} part all necessary head movements are covered. In the \textbf{Head Adjustment} part all necessary head movements are covered.
In this part of the algorithm the head is rotated by a calculated angle, which In this part of the algorithm the head is rotated by a calculated angle, which
depends on the ball yaw angle, which was provided by the \textbf{Ball depends on the ball yaw angle, which was provided by the \textbf{Ball
Detection} part. Therefore, the ball should now be aligned in the center of Detection} part. Therefore, the ball should now be aligned in the centre of
the robots' camera frames. If the angle between the head and the rest of the the robots' camera frames. If the angle between the head and the rest of the
body is now below a specified threshold, the ball is locked and the algorithm body is now below a specified threshold, the ball is locked and the algorithm
stops, otherwise the algorithm continues with \textbf{Body Adjustment}. stops, otherwise the algorithm continues with \textbf{Body Adjustment}.
@@ -56,7 +56,7 @@ current movement. Then the robot starts to rotate around its z-axis depending
on the current head yaw angle. To ensure that the head and body of the robot on the current head yaw angle. To ensure that the head and body of the robot
are aligned, like in the beginning of the whole algorithm, the head is rotated are aligned, like in the beginning of the whole algorithm, the head is rotated
back into zero yaw. The algorithm continues then with another \textbf{Ball back into zero yaw. The algorithm continues then with another \textbf{Ball
Detection}, to ensure that the robot is properly centered at the ball. Detection}, to ensure that the robot is properly centred at the ball.
The proposed algorithm provided decent results during many test runs. It allows The proposed algorithm provided decent results during many test runs. It allows
the robot to align itself to the ball fast, while some strategies are in place the robot to align itself to the ball fast, while some strategies are in place

22
documentation/perception.tex
View File

@@ -11,8 +11,8 @@ image itself needs to be transformed into HSV colorspace, so that the regions
of interest can be extracted into a \textit{binary mask}. The contours of the of interest can be extracted into a \textit{binary mask}. The contours of the
regions can then be identified in a mask \cite{contours}, and the areas of the regions can then be identified in a mask \cite{contours}, and the areas of the
regions can be calculated using the routines from the OpenCV library. The regions can be calculated using the routines from the OpenCV library. The
center and the radius of the region with the largest area are then determined centre and the radius of the region with the largest area are then determined
and are assumed to be the center and the radius of the ball. and are assumed to be the centre and the radius of the ball.
\begin{figure}[ht] \begin{figure}[ht]
\includegraphics[width=\textwidth]{\fig ball-detection} \includegraphics[width=\textwidth]{\fig ball-detection}
@@ -30,13 +30,13 @@ binary mask with erosions and dilations, which allowed us to detect the ball
even over long distances. even over long distances.
The advantages of the presented algorithm are its speed and simplicity. The The advantages of the presented algorithm are its speed and simplicity. The
major downside is that the careful color calibration is required for the major downside is that a careful color calibration is required for the
algorithm to function properly. If the HSV interval of the targeted color is algorithm to function properly. If the HSV interval of the targeted color is
too narrow, then the algorithm might miss the ball; if the interval is too too narrow, the algorithm might miss the ball; if the interval is too
wide, then other big red-shaded objects in the camera image will be detected as wide, other big red-shaded objects in the camera image will be detected as
the ball. A possible approach to alleviate these issues to a certain degree the ball. A possible approach to alleviate these issues to a certain degree
will be presented further in the section \ref{p sec field detect}. To will be presented further in the section \ref{p sec field detect}. To
conclude, we found this algorithm to be robust enough for our purposes, if the conclude, we found this algorithm to be robust enough for our purposes, if a
sensible color calibration was provided. sensible color calibration was provided.
\section{Goal Detection} \section{Goal Detection}
@@ -64,7 +64,7 @@ contours with the largest areas are considered further (in our experiments it
was empirically determined that $N=5$ provides good results). Furthermore, all was empirically determined that $N=5$ provides good results). Furthermore, all
convex contours are rejected, since the goal is a highly non-convex shape. convex contours are rejected, since the goal is a highly non-convex shape.
After that, a check is performed, how many points are necessary to approximate After that, a check is performed, how many points are necessary to approximate
the remaining contours. The motivation behind this is the following: it is the remaining contours. The motivation behind this is the following: It is
clearly visible that the goal shape can be perfectly approximated by a line clearly visible that the goal shape can be perfectly approximated by a line
with 8 straight segments. On an image from the camera, the approximation is with 8 straight segments. On an image from the camera, the approximation is
almost perfect when using only 6 line segments, and in some degenerate cases almost perfect when using only 6 line segments, and in some degenerate cases
@@ -74,7 +74,7 @@ of line segments to be approximated is probably not the goal. The preselection
stage ends here, and the remaining candidates are passed to the scoring stage ends here, and the remaining candidates are passed to the scoring
function. function.
The scoring function calculates, how different are the properties of the The scoring function calculates, how different the properties of the
candidates are from the properties, that an idealized goal contour is expected candidates are from the properties, that an idealized goal contour is expected
to have. The evaluation is happening based on two properties. The first to have. The evaluation is happening based on two properties. The first
property is based on the observation, that the area of the goal contour is much property is based on the observation, that the area of the goal contour is much
@@ -90,7 +90,7 @@ scoring function can then look like the following:
The contour, that minimizes the scoring function, while keeping its value under The contour, that minimizes the scoring function, while keeping its value under
a certain threshold is considered the goal. If no contour scores below the a certain threshold is considered the goal. If no contour scores below the
threshold, then the algorithm assumes that no goal was found. An important note threshold, the algorithm assumes that no goal was found. An important note
is that the algorithm is designed in such a way, that the preselection and is that the algorithm is designed in such a way, that the preselection and
scoring are modular, which means that the current simple scoring function can scoring are modular, which means that the current simple scoring function can
later be replaced by a function with a better heuristic, or even by some later be replaced by a function with a better heuristic, or even by some
@@ -104,7 +104,7 @@ Figure \ref{p figure goal-detection} demonstrates the algorithm in action. On
the right is the binary mask with all found contours. On the left are the goal, the right is the binary mask with all found contours. On the left are the goal,
and one contour that passed preselection but was rejected during scoring. and one contour that passed preselection but was rejected during scoring.
One downside of this algorithm, is that in some cases the field lines One downside of this algorithm is that in some cases the field lines
might appear to have the same properties, that the goal contour is expected to might appear to have the same properties, that the goal contour is expected to
have, therefore the field lines can be mistaken for the goal. We will describe, have, therefore the field lines can be mistaken for the goal. We will describe,
how we dealt with this problem, in the section \ref{p sec field detect}. how we dealt with this problem, in the section \ref{p sec field detect}.
@@ -129,7 +129,7 @@ objects on the field are properly consumed.
Such rather simple field detection has two important applications. The first Such rather simple field detection has two important applications. The first
one is that the robot should be aware, where the field is, so that it doesn't one is that the robot should be aware, where the field is, so that it doesn't
try to walk away from the field. Due to time constraints, we didn't implement try to walk away from the field. Due to time constraints, we didn't implement
this part of the behavior. The second application of field detection is the this part of the behaviour. The second application of field detection is the
improvement of the quality of goal and ball recognition. As was mentioned in improvement of the quality of goal and ball recognition. As was mentioned in
the section on ball detection, the current algorithm might get confused, if the section on ball detection, the current algorithm might get confused, if
there are any red objects in the robot's field of view. However, there there are any red objects in the robot's field of view. However, there

22
documentation/solintro.tex
View File

@@ -5,29 +5,29 @@ completed, which are:
\begin{enumerate} \begin{enumerate}
\item Ball detection; \item Ball detection
\item Goal detection; \item Goal detection
\item Field detection; \item Field detection
\item Turning to ball; \item Turning to ball
\item Distance measurement; \item Distance measurement
\item Approach planning; \item Approach planning
\item Ball approach; \item Ball approach
\item Goal alignment; \item Goal alignment
\item Ball alignment; \item Ball alignment
\item Kick. \item Kick
\end{enumerate} \end{enumerate}
In this chapter we will give our solutions to the problems posed by each of the In this chapter we will give our solutions to the problems posed by each of the
milestones, and at the end the resulting goal scoring strategy will be milestones, and at the end the resulting goal scoring strategy will be
presented. We will now start with the lower level detection milestones and presented. We will now start with the lower level detection milestones and
will gradually introduce higher level behaviors. will gradually introduce higher level behaviours.

11
documentation/tools.tex
View File

@@ -56,15 +56,14 @@ to ensure their proper functioning on the robot.
\section{Rejected Software Alternatives} \section{Rejected Software Alternatives}
Here we will briefly discuss what alternative options were available for the Here we will briefly discuss what alternative options were available for the
choice of the base framework, and why we decided not to use those. One choice of the base framework and why we decided not to use those. One
available option was the official \textit{NAOqi C++ SDK}. Being based on the available option was the official \textit{NAOqi C++ SDK}. Being based on the
C++ language, this SDK can naturally be expected to have better performance and C++ language, this SDK can naturally be expected to have better performance and to be more resource-efficient, than the Python-based version. We still chose the
be more resource-efficient, than the Python-based version. We still chose the
Python SDK, because C++ is not particularly suitable for fast prototyping, Python SDK, because C++ is not particularly suitable for fast prototyping,
because of the complexity of the language. It is also worth noting, that we because of the complexity of the language. It is also worth noting, that we
never really hit the performance constraints, that couldn't have been overcome never really hit the performance constraints, that couldn't have been overcome
by refactoring our code, but in the future it might be reasonable to migrate by refactoring our code, but in the future it might be reasonable to migrate
some of the portions of it to C++. some portions of it to C++.
Another big alternative is \textit{ROS} \cite{ros} (Robotic Operating System). Another big alternative is \textit{ROS} \cite{ros} (Robotic Operating System).
ROS is a collection of software targeted at robot development, and there exists ROS is a collection of software targeted at robot development, and there exists
@@ -79,7 +78,7 @@ in ROS, it should be possible in the future to incorporate our work into ROS.
Finally, as was already mentioned in the introduction, \textit{B-Human Finally, as was already mentioned in the introduction, \textit{B-Human
Framework} is a popular choice for beginners, thanks to the quality of the Framework} is a popular choice for beginners, thanks to the quality of the
algorithms and good documentation. However, B-Human has been in development algorithms and good documentation. However, B-Human has been in development
over many years and is therefore a very complex system. The amount time needed over many years and is therefore a very complex system. The amount of time needed
to get familiar with the code, and then to incorporate our changes would have to get familiar with the code, and then to incorporate our changes would have
been too big, for this reason we decided to use the simpler option as a been too big. For this reason we decided to use the simpler option as a
starting point. starting point.