more explanations

Makefile

@@ -40,9 +40,10 @@ RIVERS = \
 		 salvis-visvalingam-64-chaikin-50k \
 		 salvis-overlaid-douglas-64-chaikin-50k \
 		 salvis-overlaid-visvalingam-64-chaikin-50k \
-		 salvis-wm-50k \
 		 salvis-wm-250k \
-		 salvis-wm-overlaid-50k-zoom \
+		 salvis-wm-50k \
+		 salvis-wm-50k-nw \
+		 salvis-wm-50k-ne \
 		 salvis-wm-overlaid-250k-zoom \
 		 salvis-gdr50-overlaid
 
@@ -199,15 +200,22 @@ salvis-overlaid-visvalingam-64-chaikin-50k_1COLOR = orange
 salvis-overlaid-visvalingam-64-chaikin-50k_WIDTHDIV = 2
 salvis-overlaid-visvalingam-64-chaikin-50k_QUADRANT = 1
 
-salvis-wm-50k_1SELECT = wm_visuals where name='salvis-wm-75'
-salvis-wm-50k_WIDTHDIV = 2
-
 salvis-wm-250k_1SELECT = wm_visuals where name='salvis-wm-220'
 salvis-wm-250k_WIDTHDIV = 10
 
-salvis-wm-overlaid-50k-zoom_1SELECT = wm_visuals where name='salvis-wm-75'
+salvis-wm-50k_1SELECT = wm_visuals where name='salvis-wm-75'
-salvis-wm-overlaid-50k-zoom_2SELECT = wm_visuals where name='salvis'
+salvis-wm-50k_2SELECT = wm_visuals where name='salvis'
-salvis-wm-overlaid-50k-zoom_1COLOR = orange
+salvis-wm-50k_1COLOR = orange
+
+salvis-wm-50k-nw_1SELECT = wm_visuals where name='salvis-wm-75'
+salvis-wm-50k-nw_2SELECT = wm_visuals where name='salvis'
+salvis-wm-50k-nw_1COLOR = orange
+salvis-wm-50k-nw_QUADRANT = 2
+
+salvis-wm-50k-ne_1SELECT = wm_visuals where name='salvis-wm-75'
+salvis-wm-50k-ne_2SELECT = wm_visuals where name='salvis'
+salvis-wm-50k-ne_1COLOR = orange
+salvis-wm-50k-ne_QUADRANT = 1
 
 salvis-wm-overlaid-250k-zoom_1SELECT = wm_visuals where name='salvis-wm-220'
 salvis-wm-overlaid-250k-zoom_2SELECT = wm_visuals where name='salvis'

mj-msc.tex

@@ -1378,45 +1378,49 @@ table~\ref{table:scale-halfcirlce-diameter}:
 \subsubsection{Medium-scale (1:\numprint{50000})}
 \label{sec:analyzed-medium-scale}
 
-\begin{figure}[ht]
+\begin{figure}[h!]
-    \centering
-    \begin{subfigure}[b]{.49\textwidth}
-        \includegraphics[width=\textwidth]{salvis-50k}
-        \caption{Original.}
-    \end{subfigure}
-    \hfill
-    \begin{subfigure}[b]{.49\textwidth}
     \centering
     \includegraphics[width=\textwidth]{salvis-wm-50k}
-        \caption{Simplified.}
+    \caption{2x zoomed-in {\WM} for 1:\numprint{50000}.}
-    \end{subfigure}
-    \caption{Simplified with {\WM} for 1:\numprint{50000}.}
     \label{fig:salvis-wm-50k}
 \end{figure}
 
 As one can see in figure~\ref{fig:salvis-wm-50k}, the illustrations deliver
-what was promised by the algorithm, but with a few caveats. West-side of the
+what was promised by the algorithm, but with a few caveats. Left side of the
-river looks reasonably well simplified: long bends remain slightly curved,
+figure looks reasonably well simplified: long bends remain slightly curved,
 small bends are removed or slightly exaggerated.
 
-Approximately at the middle of the figure some jagged and sharp bends appear.
+Figure's~\ref{fig:salvis-wm-50k} left part is clipped to
-Figure~\ref{fig:salvis-wm-50k-overlaid-zoom} zooms in, overlaying the original
+figure~\ref{fig:salvis-wm-50k-nw}. As one can see, some bends were well
-river (in orange). These will become more pronounced in even larger-scale
+exaggerated, and some bends were eliminated.
+
+\begin{figure}[h!]
+    \centering
+    \includegraphics[width=\textwidth]{salvis-wm-50k-nw}
+    \caption{Left part of figure~\ref{fig:salvis-wm-50k}.}
+    \label{fig:salvis-wm-50k-nw}
+\end{figure}
+
+Top--right side (clipped in figure~\ref{fig:salvis-wm-50k-ne}) some jagged
+and sharp bends appear. These will become more pronounced in even larger-scale
 simplification in the next section.
 
-\begin{figure}[ht]
+\begin{figure}[h!]
     \centering
-    \includegraphics[width=\textwidth]{salvis-wm-overlaid-50k-zoom}
+    \includegraphics[width=\textwidth]{salvis-wm-50k-ne}
-    \caption{2x zoomed-in {\WM} for 1:\numprint{50000}.}
+    \caption{Top--right part of figure~\ref{fig:salvis-wm-50k}.}
-    \label{fig:salvis-wm-50k-overlaid-zoom}
+    \label{fig:salvis-wm-50k-ne}
 \end{figure}
 
+To sum up, mid-scale simplification works well for some geometries, but creates
+sharp edges for others.
+
 \subsubsection{Large-scale (1:\numprint{250000})}
 
 As visible in figure~\ref{fig:salvis-wm-250k}, for large-scale map, some of the
 resulting bends look significantly exaggerated. Why is that?
 Figure~\ref{fig:salvis-wm-250k-overlaid-zoom} zooms in the large-scale
-simplification and overlays the original (in orange).
+simplification and overlays the original.
 
 \begin{figure}[ht]
     \centering
@@ -1431,7 +1435,7 @@ simplification and overlays the original (in orange).
         \includegraphics[width=.2\textwidth]{salvis-wm-250k}
         \caption{Simplified.}
     \end{subfigure}
-    \caption{Simplified with {\WM} for 1:\numprint{250000}.}
+    \caption{GDB10LT simplified with {\WM} for 1:\numprint{250000}.}
     \label{fig:salvis-wm-250k}
 \end{figure}
 
@@ -1442,11 +1446,27 @@ simplification and overlays the original (in orange).
     \label{fig:salvis-wm-250k-overlaid-zoom}
 \end{figure}
 
-A conglomeration of bends is visible, especially in north-east side of the
+A conglomeration of bends is visible, especially in top--right side of the
-center. We assume this was caused by two bends significantly exaggerated,
+illustration. We assume this was caused by two bends significantly exaggerated,
-leaving no space to exaggerate for those between the two. For the same reason,
+leaving no space to exaggerate for those between the two.
-the figure~\ref{fig:salvis-wm-50k} (of scale 1:\numprint{50000}) has many
+
-smaller bends at approximately the same location.
+\subsubsection{Discussion}
+
+For mid-size scales of 1:\numprint{50000}, the implemented algorithm works well
+for certain geometries, and poorly for others. This test surfaced two areas for
+future research and improvement:
+
+\begin{itemize}
+
+    \item Exaggeration is sometimes creating sharp edges, especially when the
+        exaggerated bend is quite small. When sharp edges are created,
+        exaggeration could interpolate more points in the bend, and exaggerate
+        using the interpolated points.
+
+    \item In larger scales, when bends do not have space to exaggerate, they
+        should be combined or eliminated instead.
+
+\end{itemize}
 
 \subsection{Comparing Generalized Result with National Spatial Datasets}
 
@@ -1547,7 +1567,7 @@ implementation:
 
     \item Fine-tune parameters for bend exaggeration.
         Section~\ref{sec:generalization-results-of-analyzed-rivers} contains
-        a exaggerated bends that became sharp.
+        a exaggerated bends that became sharp and includes some future ideas.
 
     \item Research when bends should be marked as \textsc{isolated}. As is
         seen from examples, the current criteria is not robust enough.