A Novel Parallel Algorithm with Map Segmentation for Multiple Geographical Feature Label Placement Problem

Multiple geographical feature label placement (MGFLP) is an NP-hard problem that can negatively influence label position accuracy and the computational time of the algorithm. The complexity of such a problem is compounded as the number of features for labeling increases, causing the execution time of the algorithms to grow exponentially. Additionally, in large-scale solutions, the algorithm possibly gets trapped in local minima, which imposes significant challenges in automatic label placement. To address the mentioned challenges, this paper proposes a novel parallel algorithm with the concept of map segmentation which decomposes the problem of multiple geographical feature label placement (MGFLP) to achieve a more intuitive solution. Parallel computing is then utilized to handle each decomposed problem simultaneously on a separate central processing unit (CPU) to speed up the process of label placement. The optimization component of the proposed algorithm is designed based on the hybrid of discrete differential evolution and genetic algorithms. Our results based on real-world datasets confirm the usability and scalability of the algorithm and illustrate its excellent performance. Moreover, the algorithm gained superlinear speedup compared to the previous studies that applied this hybrid algorithm.

Optimal layout of stacked graph for visualizing multidimensional financial time series data

In the era of big data, the analysis of multi-dimensional time series data is one of the important topics in many fields such as finance, science, logistics, and engineering. Using stacked graphs for visual analysis helps to visually reveal the changing characteristics of each dimension over time. In order to present visually appealing and easy-to-read stacked graphs, this paper constructs the minimum cumulative variance rule to determine the stacking order of each dimension, as well as adopts the width priority principle and the color complementary principle to determine the label placement positioning and text coloring. In addition, a color matching method is recommended by user study. The proposed optimal visual layout algorithm is applied to the visual analysis of actual multidimensional financial time series data, and as a result, vividly reveals the characteristics of the flow of securities trading funds between sectors.

Implementation and Evaluation of a Fast Area Feature Labeling Method Using Auxiliary Lines

Appropriate place labels, which provide the name or attribute of a graphical feature, are important in geographical information systems and cartography. Herein, an internal label placement method was proposed for area features, such as cities, prefectures, and lakes, on a map. For internal label placement, placing a large label for an extremely narrow or small area, such that the label does not protrude from the corresponding area is challenging. In such cases, a label can overlap with protruding labels from other areas. Meanwhile, tablet devices have been rapidly employed in recent years. Because tablet devices can easily zoom in on a map, it is possible to eliminate the overlaps by enlarging the map without changing the label size. Therefore, we proposed a method that enables real-time processing, even on tablet devices. The label positions are determined by detecting the intersections of the auxiliary and boundary lines of a given area feature. The proposed method adequately labels the positions of area features, even those with indents and narrow sections. Moreover, it can find tens of thousands of label positions within 100 ms, even on low-performance computers, such as tablet devices.

AUTOMATIC LABEL PLACEMENT OF AREA-FEATURES USING DEEP LEARNING

Label placement is one of the most essential tasks in the fields of cartography and geographic information systems. Numerous studies have been conducted on the automatic label placement for the past few decades. In this study, we focus on automatic label placement of area-feature, which has been relatively less studied than that of point-feature and line-feature. Most of the existing approaches have adopted a rule-based algorithm, and there are limitations in expressing the characteristics of label placement for area-features of various shapes utilizing handcrafted rules, criteria, objective functions, etc. Hence, we propose a novel approach for automatic label placement of area-feature based on deep learning. The aim of the proposed approach is to obtain the complex and implicit characteristics of area-feature label placement by manual operation directly and automatically from training data. First, the area-features with vector format are converted into a binary image. Then a key-point detection model, which simultaneously detect and localize specific key-points from an image, is applied to the binary image to estimate the candidate positions of labels. Finally, the final label placement positions for each area-feature are determined via simple post-process. To evaluate the proposed approach, the experiments with cadastral data were conducted. The experimental results show that the ratios of the estimation errors within 1.2 m (corresponding to one pixel of the input image) were 92.6% and 94.5% in the center and upper-left placement style, respectively. It implies that the proposed approach could place the labels for area-features automatically and accurately.

Using artificial neural network for labelling polygon features in topographic maps

The purpose of this article was to present the methodology which enables automatic map labelling. This topic is particularly important in the context of the ongoing research into the full automation of visualization process of spatial data stored in the currently used topographic databases (e.g. OpenStreetMap, Vector Map Level 2, etc.). To carry out this task, the artificial neural network (multilayer perceptron) was used. The Vector Map Level 2 was used as a test database. The data for neural network learning (the reference label localization) was obtained from the military topographic map at scale 1 : 50 000. In the article, the method of applying artificial neural networks to the map labelling is presented. Detailed research was carried out on the basis of labels from the feature class “built-up area”. The results of the analyses revealed that it is possible to use the artificial intelligence computational methods to automate the process of placing labels on maps. The results showed that 65% of the labels were put on the topographic map in the same place as in the case of the labelling which was done manually by a cartographer. The obtained results can contribute both to the enhancement of the quality of cartographic visualization (e.g. in geoportals) and the partial elimination of the human factor in this process. Highlights for public administration, management and planning: • Map label placement is among key variables ensuring the usability of topographic maps across disciplines. • We present the neural network approach for automating the process of labelling topographic maps with locality names. • The presented case study applies to the military map in scale 1:50 000, but can be applied on other maps and geoportals.

Point Label Placement on Hexagonal Map Grids

<p><strong>Abstract.</strong> The past few decades have seen the development of automatically feature labelling when manual label placement was thought to be time and labour consuming. Emerging techniques like volunteered geographic information (VGI) collection are making label placement more complexed with many features in a limited space, especially for points of interest (POI). In order to improve the quality and the efficiency of point feature labelling, there have been massive researches focusing on issues like position models, assessment criteria and optimization methods. Most of the researches were using vector-based methods while raster-based methods were less used, because vector-based methods have the advantage of easy definition of features and labels but are usually followed by computation complexity problems for features with high density. In contrast raster-based methods are faster and more flexible, though being harder to represent features and labels precisely on the map grids. Considering that hexagon partitioning was rarely used in raster-based methods, compared with the most commonly used square portioning, and hexagon was potentially useful for its oblique sides and isotropic orientations, hexagonal grids were used in this research to investigate better point feature labelling approaches.</p><p>A new raster-based method was promoted to figure out high quality label placement of POI in dense area. Labels were placed on a hexagonal map grids based on the principles that one Chinese character is set to one hexagon unit with the mathematical relationship of <i>h</i>&amp;thinsp;=&amp;thinsp;((&amp;radic;3+1)/2)<i>a</i>, while <i>h</i> is the side length of a hexagon unit and <i>a</i> is the size of a Chinese character. Considering that hexagon grids are divided into flat topped type and pointy topped type, which leads to different orientations, split hexagons were promoted to extend orientations from 6 to 8 based on pointy topped grids. A hexagon is partitioned into two parts labelled ‘left’ and ‘right’ and a split hexagon is the combination of a ‘left’ part and a ‘right’ part separately from two neighboring hexagons, as shown in figure 1. Then every hexagon on the grid will have four status: not-occupied {(0,0)}, half-occupied {(0,1) and (1,0)} and both-occupied {(1,1)}. Based on the fundamental concepts above, specific definitions were made on how labels were supposed to be represented on hexagonal map grids, including the length, orientation, writing direction, character orientation and position of the labels.</p><p>The approach first initially arranges labels of POI with different combinations of label orientations while pursuing coherence as much as possible, including procedures of rasterization of vector data, POI grouping and initial scheme computation. Every POI in a same group would have same label orientation and every POI group may have several accessible orientations thus making initial schemes diverse. Then a second positioning algorithm was conducted to handle overlapping (labels with POI, labels with labels) problems and improve the overall quality of labelling. The algorithm used the methods of position changing and label turning, which allow label to change its position around POI and sometimes change the orientation when it is necessary to avoid collisions. Quality of labels in a closed block was assessed from three aspects: preferential orientation, occlusion and spaciousness. POI data was chosen from restaurant, hotel and shop facilities and figure 2 showed one of the examples of label placement results using this method. The results have shown good orientation consistency of labels and occlusions were reduced to the lowest, though several label-label occlusions remained due to the limited space. After being compared with vector-based method, the approach has shown better performance on maintaining map legibility, aesthetics and harmony.</p>