我需要你的帮助，才能让这两个长表格看起来更好。另外，如果你能对可能用到的软件包和命令发表评论，以便了解它们的工作原理，我将不胜感激。
表格如下：
代号：T5
\begin{table}[htb]
\caption{The state-of-the-art in taxonomy, detection, extraction and pattern recognition in urban applications}
\label{T2.5}
\centering
\resizebox{\columnwidth}{!}{%
\begin{tabular}{| l | l | l | l | l |}
\hline
\textbf{Application} & \textbf{Taxonomy/quantification algorithms} & \textbf{Merits} & \textbf{Limitations} & \textbf{Example} \\
\hline
\multirow{5}{*}{} Seismic building structural types (SBSTs) & Support vector machine (SVM) and random forest (RF) & Classify the combination of different remote sensing data & Hierarchical supervised classification scheme has uncertainties in separating SBSTs & \cite{geiss2015} \\ \cline{2-5}
& & Derive sets of valuable features to characterise the urban environment & Performance depends on the ranked features & \\ \cline{2-5}
& & Model an effective earthquake loss and spatial distribution & Accuracy is subject to the addition of further features and subset based categories & \\ \cline{2-5}
& & ABTSVM classifier outperforms other multi-class SVM classifiers & & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Urban change detection: landuse transitions & Scene classification with a bag-of-visual-words (BOVW) & Obtain semantic scene classes & Time-consuming & \cite{wu2016} \\ \cline{2-5}
& & Effectively analyse landuse changes & Very difficult to achieve the direct selection of the “from-to” samples from the dataset & \\ \cline{2-5}
& & Satisfactory accuracy & Classification performance in certain cases is negatively affected by the redundancy of information & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Urban landcover and landuse classification & Binary tree SVM based on Jeffries–Matusita (JM) distance & Improve classification accuracy & Confusion between road and bare soil classes & \cite{du2012} \\ \cline{2-5}
& & Classify hyperspectral images adequately & Instability and complexity in the structure and parameters of the binary tree SVM & \\ \cline{2-5}
& & Ease of interpretation of urban classes & & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Monitoring changes in impervious surfaces & MRGU & Integrate multiple endmember spectral mixture analysis (MESMA) & Not universally applicable due to specific behaviour of maximum noise fraction (MNF) & \cite{shahtahmassebi2016} \\ \cline{2-5}
& & Analysis of regression residuals & Difficult to use for quantifying changes in urban centres & \\ \cline{2-5}
& & Spatial statistics (Getis Ord) with Moran’s I & Performance (regression residuals) is subject to the structure of urban regions & \\ \cline{2-5}
& & Urban growth theories in an effective manner & & \\ \cline{2-5}
& & Quantify the magnitude of impervious surface changes & & \\ \cline{2-5}
& & Identify spatial distribution& & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Change detection & Unsupervised Neural network and feature transformation & SDAE for better representation of the relationships between feature- and pixel-pair & Uncertainty in the feature-pair & \cite{zhang2016} \\ \cline{2-5}
& & Mapping based FCA and function learning to highlight change in a robust manner & High computational cost & \\ \cline{2-5}
& & Denoising autoencoder & Complexity structures of stacked denoising autoencoders (SDAE) & \\ \cline{2-5}
& & & Many constraints are required to extract useful features & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Water quality, and sustainable water resources management & Integrated data fusion and mining (IDFM) and artificial neural network (ANN) & A near real-time monitoring and the early warning system & Prediction accuracy may be effected by uncertainties in the fused data & \cite{imen2015} \\ \cline{2-5}
& & Efficiency & A large number of variables is required to overcome the uncertainty & \\ \cline{2-5}
& & Forecasting reliability & Not applicable for regional meteorology parameters & \\ \cline{2-5}
& & Potential for local adoption & & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Air temperature estimation & SVM & Fully automated method & Requires expert users to apply SVM & \cite{moser2015} \\ \cline{2-5}
& & SVM regression is robust & Does not work well under non cloud-free conditions and require in situ measurements & \\ \cline{2-5}
& & Regression errors can be modelled at the pixel level, improving accuracy estimation & Regression error distribution is insufficient & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Fine-scale population estimation for urban management, emergency response and epidemiological & RF and Linear regression modelling & Able to classify building types and extract their footprints in the heterogeneous urban areas & Subject to the accuracy of the selected morphology filter & \cite{xie2015} \\ \cline{2-5}
& & Improved classification accuracy & Use of large numbers of metrics and variables for building type classification & \\ \cline{2-5}
& & Ease of adoption & Building background metrics do not show its advantage in the block classification & \\ \cline{2-5}
& & & Classification uncertainty for non-residential buildings & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Renewable energy and urban feature extraction & Shadow detection and building geometry identification & Easy to apply & Semi-automated & \cite{kadhim2015a} \\ \cline{2-5}
& & Sufficient to generate 3D model of urban buildings & Not suitable for dense urban areas & \\ \cline{2-5}
& & Reliable analysis of the solar energy potential & Sensitive to the quality of satellite images & \\ \cline{2-5}
& & Identify the availability of 3D surfaces & & \\ \cline{2-5}
& & Flexibility and feasibility & & \\ \cline{2-5}
\hline
\multirow{5}{*}{} Impervious surfaces estimation & SVM and RF & Increased classification accuracy & Needs many texture matrices & \cite{zhang2015} \\ \cline{2-5}
& & Sufficient to generate 3D model of urban buildings & Not suitable for dense urban areas & \\ \cline{2-5}
& & Does not depend on combinations of features & Inability to handle the confusion in shaded areas and bare soil & \\ \cline{2-5}
& & Data can be fused to optimise parameters efficiently & Over-fitting & \\ \cline{2-5}
& & Ease of application & & \\ \cline{2-5}
\hline
\end{tabular}
}
\end{table}
第二个，T6 是：
\begin{table}[]
\centering
\caption{Directions of future research}
\label{T2.6}
\resizebox{\columnwidth}{!}{%
\begin{tabular}{l|l|l|l|l|l|l|}
\cline{2-7}
& \multicolumn{2}{l|}{\textbf{Integration of heterogeneous data}} & \multicolumn{2}{l|}{\textbf{Algorithms to identify urban features}} & \multicolumn{2}{l|}{\textbf{Improve accuracy for spectral classification algorithms}} \\ \hline
\multicolumn{1}{|l|}{\textbf{Objective}} & \multicolumn{2}{l|}{\begin{tabular}[c]{@{}l@{}} Improve the spatial and spectral resolution\\ Enhance the ability of features detection and display\\ Promote the geometric precision\\ Increase the capability of the change detection\\ Refine, replace or repair the defect of image data\\ Refine, replace or repair the defect of image data\\ Handle multi-source remote sensing data at the pixel, feature and decision of level fusion\end{tabular}} & \multicolumn{2}{l|}{\begin{tabular}[c]{@{}l@{}} Accelerate future processing and improve classification accuracy\\ Automated processes for detecting, extracting, simulating, classifying and modelling urban features\\ Capability of handling and fusing the large number of datasets\\ Improve image objects segmentation\\ Increase the reliability and precision of the feature extraction\\ Mitigate the ambiguous and uncertainty\end{tabular}} & \multicolumn{2}{l|}{\begin{tabular}[c]{@{}l@{}} Capability of separating urban land-cover and land-use classes in an adequate manner\\ For a better visualisation and interpretation of urban landscape\\ Performing a change detection analysis and pattern recognition\\ Accurately characterise the model parameters for different urban remote sensing applications\end{tabular}} \\ \hline
\multicolumn{1}{|l|}{sfre} & \multicolumn{2}{l|}{\begin{tabular}[c]{@{}l@{}}A rescaling of multisource data of divers EO instruments\\Assessing the distortion of the spatial and spectral resolution\\ Time-consuming and subjective\\ The complexity and/or availability of a large training datasets for the deep learning features\\ Manual or semi-automated the post-processing process\\ Computation efficiency and effectiveness\\ The quality of the distinguishing features\\ With the fusion schemes, the optimal combining strategy of the current fusion algorithms is a challenging task and promising research that requires further investigations in the near future
\end{tabular}} & \multicolumn{2}{l|}{\begin{tabular}[c]{@{}l@{}}Further developments for fusing LiDAR data with thermal, multispectral and hyperspectral imagery\\ Reliable determination of the boundaries of urban objects in an automated manner\\ Further improvements for addressing different characteristics of EO data\\Ability to cope with unpredictable environmental and illumination factors of diverse datasets\\ Similarity in the characteristics of spectral, textural and geometrical-based between the urban objects and their background\\ Similarity in the characteristics of spectral, textural and geometrical-based between the urban objects and their background\\ Computational time to perform a task\end{tabular}} & \multicolumn{2}{l|}{\begin{tabular}[c]{@{}l@{}}Mixed pixels\\ Uncertainty in the urban land-cover and land-use classes\\Mixed objects\\Promote pixel-based and object-based classification using contextual information\\ Over-fitting can cause speckled results that are difficult to interpret\\ An automatic labelling strategy is required for actual label sets in several applications\\ Further refinements for the fusion of diverse data sources\\ Integrate the derived urban features (e.g. building shadows) and/or spatial matrices with various classifier schemes\\ Investigate the recent computer vision techniques for improving the accuracy \end{tabular}} \\ \hline
\end{tabular}
}
\end{table}
我将非常感激任何可能给予我的帮助。
这是我的论文文档代码的开头：
%% ----------------------------------------------------------------
%% Thesis.tex -- MAIN FILE (the one that you compile with LaTeX)
%% ----------------------------------------------------------------
% Set up the document
\documentclass[a4paper, 11pt, arial, oneside]{Thesis} % Use the "Thesis" style, based on the ECS Thesis style by Steve Gunn
\usepackage{charter}
\usepackage[scaled=0.92]{helvet}
%\usepackage [T1] {fontenc}
%\renewcommand\familydefault{phv}
答案1
只是采纳我对你上一个问题的回答（如何使用 Latex 让这张表看起来更好）用于表格数据，并修改表格主体米科答案......并添加新的数据列表外观-)：
\documentclass[11pt,a4paper]{report}
\usepackage{charter}
\usepackage[scaled=0.92]{helvet}
\usepackage[margin=2.5cm]{geometry} % set width of textblock appropriately
\usepackage[english]{babel}
\usepackage{ragged2e}
\usepackage{array, booktabs, tabularx}
\newcolumntype{L}{>{\RaggedRight\hspace{0pt}}X}
\usepackage[skip=1ex,labelfont=bf,font=small]{caption}
\usepackage{enumitem}
\usepackage{etoolbox}
\AtBeginEnvironment{table}{%
\setlist[itemize]{nosep,
wide,%leftmargin = *,
before = \vspace{-\baselineskip},
after = \vspace{-\baselineskip}
}
}% end of AtBeginEnvironment
%-------------------------------- show page layout, only for test
\usepackage{showframe}
\renewcommand\ShowFrameLinethickness{0.15pt}
\renewcommand*\ShowFrameColor{\color{red}}
%---------------------------------------------------------------%
\begin{document}
\setcounter{chapter}{1} % just for this example
\setcounter{table}{5} % just for this example
\begin{table}[htbp]
\small
\setlength\tabcolsep{2pt} % default: 6pt
\caption{Directions for future research} \label{T2.6}
%
\begin{tabularx}{\textwidth}{@{}p{9ex} L L L}
\toprule
&
Integration of heterogeneous data &
Algorithms to identify urban features &
Improve accuracy for spectral classification algorithms \\
\midrule
Objective &
\begin{itemize}
\item Improve the spatial and spectral resolution
\item Enhance the ability of features detection and display
\item Promote the geometric precision
\item Increase the capability of the change detection
\item Refine, replace or repair the defect of image data
\item Refine, replace or repair the defect of image data
\item Handle multi-source remote sensing data at the pixel, feature and
decision of level fusion
\end{itemize}
&
\begin{itemize}
\item Accelerate future processing and improve classification accuracy
\item Automated processes for detecting, extracting, simulating, classifying and modelling urban features
\item Capability of handling and fusing the large number of datasets
\item Improve image objects segmentation
\item Increase the reliability and precision of the feature extraction
\item Mitigate the ambiguous and uncertainty
\end{itemize}
&
\begin{itemize}
\item Capability of separating urban land-cover and land-use classes in an adequate manner
\item For a better visualisation and interpretation of urban landscape
\item Performing a change detection analysis and pattern recognition
\item Accurately characterise the model parameters for different urban remote sensing applications
\end{itemize} \\
\midrule
Problems requiring solutions
&
\begin{itemize}
\item A rescaling of multisource data of divers EO instruments
\item Assessing the distortion of the spatial and spectral resolution Time-consuming and subjective
\item The complexity and\slash or availability of a large training datasets for the deep learning features
\item Manual or semi-automated the post-processing process Computation efficiency and effectiveness
\item The quality of the distinguishing features
\item With the fusion schemes, the optimal combining strategy of the current fusion algorithms is a challenging task and promising research that requires further investigations in the near future
\end{itemize}
&
\begin{itemize}
\item Further developments for fusing LiDAR data with thermal, multispectral and hyperspectral imagery
\item Reliable determination of the boundaries of urban objects in an automated manner
\item Further improvements for addressing different characteristics of EO data
\item Ability to cope with unpredictable environmental and illumination factors of diverse datasets
\item Similarity in the characteristics of spectral, textural and geometrical-based between the urban objects and their background
\item Similarity in the characteristics of spectral, textural and geometrical-based between the urban objects and their background
\item Computational time to perform a task
\end{itemize}
&
\begin{itemize}
\item Mixed pixels
\item Uncertainty in the urban land-cover and land-use classes
\item Mixed objects
\item Promote pixel-based and object-based classification using contextual information
\item Over-fitting can cause speckled results that are difficult to interpret
\item An automatic labelling strategy is required for actual label sets in several applications
\item Further refinements for the fusion of diverse data sources
\item Integrate the derived urban features (e.g., building shadows) and\slash or spatial matrices with various classifier schemes
Investigate the recent computer vision techniques for improving the accuracy
\end{itemize}\\
\bottomrule
\end{tabularx}
\end{table}
\end{document}
编辑：似乎tabularx没有像我预期的那样工作。因此我改变了
\begin{tabularx}{\textwidth}{@{}>{\hsize=0.4\hsize}L *{3}{>{\hsize=1.2\hsize}L}@{}}
到
\begin{tabularx}{\linewidth}{@{}p{9ex} LLL @{}}
因此现在单元格中的文本更适合并且表格更低。
附录：
- 你的问题是基于个人观点的。不同的人有不同的品味:-)，并提出不同的解决方案，
- 然而，餐桌设计的第一条规则是始终如一。如果您决定使用某种表格样式，那么您应该在所有表格中使用它。此外，如果表格中有列表，那么所有表格都必须是相同的形式。
- 特殊的问题是表格单元格中的长文本。在这种情况下，使用
\small字体大小来表示表格内容是明智的。这样文本就可以更好地适应单元格。 - 您应该记住，乳胶期望页面上大约有 1/3 的文本。如果您的表格较长，请考虑使用
longtable或ltablex将其写在更多页面上。 - 在表格设计中建议考虑好的做法，例如可以在互联网上找到“少即是多”ETC
答案2
这是您正在处理的两个表中的第二个表（即较小的一个表）的截图。我认为现在它更清晰了。
请注意，我无法访问您的文档类。我已用通用report类替换。
\documentclass[11pt,a4paper]{report}
\usepackage{charter}
\usepackage[scaled=0.92]{helvet}
\usepackage[margin=2.5cm]{geometry} % set width of textblock appropriately
\usepackage[english]{babel}
\usepackage{graphicx,tabularx,ragged2e,booktabs}
\newcolumntype{L}{>{\RaggedRight\setlength\parskip{0.2\baselineskip}\arraybackslash}X}
\setlength\extrarowheight{1pt}
\usepackage{caption}
\newlength\mylen
\settowidth{\mylen}{Objective} % measure width of longest word in 1st col.
\begin{document}
\setcounter{chapter}{1} % just for this example
\setcounter{table}{5} % just for this example
\begin{table}[htbp]
\small
\setlength\tabcolsep{3pt} % default: 6pt
\captionsetup{font=small,skip=0.333\baselineskip}
\caption{Directions for future research} \label{T2.6}
\begin{tabularx}{\textwidth}{@{} p{\mylen} LLL @{}}
\toprule
&
Integration of heterogeneous data &
Algorithms to identify urban features &
Improve accuracy for spectral classification algorithms \\
\midrule
Objective &
Improve the spatial and spectral resolution \par
Enhance the ability of features detection and display \par
Promote the geometric precision \par
Increase the capability of the change detection \par
Refine, replace or repair the defect of image data \par
Refine, replace or repair the defect of image data \par
Handle multi-source remote sensing data at the pixel, feature and
decision of level fusion &
Accelerate future processing and improve classification accuracy\par
Automated processes for detecting, extracting, simulating, classifying and modelling urban features\par
Capability of handling and fusing the large number of datasets\par
Improve image objects segmentation\par
Increase the reliability and precision of the feature extraction\par
Mitigate the ambiguous and uncertainty &
Capability of separating urban land-cover and land-use classes in an adequate manner\par
For a better visualisation and interpretation of urban landscape\par
Performing a change detection analysis and pattern recognition\par
Accurately characterise the model parameters for different urban remote sensing applications \\
\midrule
Problems requiring solutions &
A rescaling of multisource data of divers EO instruments\par
Assessing the distortion of the spatial and spectral resolution\par Time-consuming and subjective\par
The complexity and\slash or availability of a large training datasets for the deep learning features\par
Manual or semi-automated the post-processing process\par Computation efficiency and effectiveness\par
The quality of the distinguishing features\par
With the fusion schemes, the optimal combining strategy of the current
fusion algorithms is a challenging task and promising research that
requires further investigations in the near future &
Further developments for fusing LiDAR data with thermal, multispectral and hyperspectral imagery\par
Reliable determination of the boundaries of urban objects in an automated manner\par
Further improvements for addressing different characteristics of EO data\par
Ability to cope with unpredictable environmental and illumination factors of diverse datasets\par
Similarity in the characteristics of spectral, textural and geometrical-based between the urban objects and their background\par
Computational time to perform a task &
Mixed pixels\par
Uncertainty in the urban land-cover and land-use classes\par
Mixed objects\par
Promote pixel-based and object-based classification using contextual information\par
Over-fitting can cause speckled results that are difficult to interpret\par
An automatic labelling strategy is required for actual label sets in several applications\par
Further refinements for the fusion of diverse data sources\par
Integrate the derived urban features (e.g., building shadows) and\slash or spatial matrices with various classifier schemes\par
Investigate the recent computer vision techniques for improving the accuracy
\\
\bottomrule
\end{tabularx}
\end{table}
\end{document}