I have a list and I want to display a figure after each list item. I get first two figures correctly, but not the last two. Here is my code:
\begin{enumerate}[leftmargin=0cm,itemindent=.5cm,labelwidth=\itemindent,labelsep=0cm,align=left]
\item
\textbf {Chi-squared distance :} The chi squared distance measures the distance between two histograms having identical bins (16 bins along each dimension). Moreover, both histograms are normalized, i.e. their entries sum up to one. The distance measure $\bf d$ is usually defined (although alternative definitions exist) as \[ d(H_1,H_2) = \Sigma \frac{ (H_1(I)-H_2(I))^{2} }{ (H_1(I)+H_2(I)) }\]The name of the distance is derived from Pearson's chi squared test statistic $ X^{2} (x,y) = \Sigma( (x_i-y_i)^{2} / x_i)$ for comparing discrete probability distributions (i.e histograms). However, unlike the test statistic, $d(H_1,H_2)$ is symmetric wrt. $H_1$ and $H_2$. The histograms obtained for superpixels are very sparse and only bins with non-zero entries in atleast one of the two histograms are considered. Thus, the bins with zero entries in both $H_1$ and $H_2$ histograms are discarded. An example of obtained chi-squared distance values is as follows:
\begin{figure}[!h]
\centering
\includegraphics [width=90mm]{chi_sqr.png}
\caption{Variation of chi squared distance for a superpixel over consecutive image frames}
\end{figure}
\item
\textbf{Correlation :} Here, the correlation or similarity between two histograms having identical bins is calculated. A higher value of correlation indicates more similarity and thus similar labels. If the correlation value for a superpixel drops below a certain threshold, a new label should be assigned to that superpixel. Correlation is calculated using \[ d(H_1,H_2) = \frac{ \Sigma_I (H_1(I)-\overline H_1)(H_2(I)-\overline H_2) }{\sqrt{ \Sigma_I (H_1(I)-\overline H_1)^{2}\Sigma_I (H_2(I)-\overline H_2)^{2}} }\] where \[ \overline H_k = \frac{1}{N}\Sigma_j H_k(j)\]
\begin{figure}[!h]
\centering
\includegraphics [width=90mm]{correl.png}
\caption{Correlation values for a superpixel over consecutive image frames}
\end{figure}
\item
\textbf{Mean squared distance :} MS distance is the sum of the squared differences of the histogram bins. It is computed as follows: \[ d(H_1, H_2) = \Sigma_I (H_1(I) - H_2(I))^2\]
\begin{figure}[!h]
\centering
\includegraphics [width=90mm]{mse.png}
\caption{Variation of mean squared distance for a superpixel over consecutive image frames}
\end{figure}
\item
\textbf{Average Color :} Here, the average color of the region enclosed by each superpixel is computed. For LAB input image, avergae color is caluclated for each channel (L, A, B) separately. The color difference between superpixels in different frames, $d(F_1, F_2)$ is obtained using \[d(F_1, F_2) = (L_{F_1} - L_{F_2})^2 + (A_{F_1} - A_{F_2})^2 + (B_{F_1} - B_{F_2})^2\]
\begin{figure}[!h]
\centering
\includegraphics [width=90mm]{avg_color.png}
\caption{Variation of average color difference for a superpixel over consecutive image frames}
\end{figure}
\end{enumerate}
Please tell me how to get my figures at the positions specified in the code.
floatpackage? You can use the[H]options fin figures. More info in the TeX FAQ – TeXtnik May 27 '13 at 09:45figure) when you don't want your content to float? – Martin Schröder May 27 '13 at 14:56subfigureusing captions. Describe the whole figure in the body of the document, it will be much more organized and pleasant to read. – juliohm May 27 '13 at 15:31