| Author: | Peng, Qian |
| Title: | Complexity-based optimization of cartographic design for multi-scale image-map generation |
| Advisors: | Li, Zhilin (LSGI) |
| Degree: | Ph.D. |
| Year: | 2022 |
| Subject: | Remote-sensing maps Maps -- Design Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Land Surveying and Geo-Informatics |
| Pages: | viii, 114 pages : color illustrations |
| Language: | English |
| Abstract: | Maps are graphic representations of geographical features and their spatial relationship. However, maps appear abstract to many people due to the use of geometric symbols. Satellite images, on the other hand, appear more realistic owing to their resemblance to real-world scenery. A hybrid design with the advantage of maps' high interpretation efficiency and satellite images' realism is desired, leading to image-maps. However, the quality of such map is criticized by many people. When overlaying map symbols onto images, discrepancies have been found due to the limited geometric accuracy of map elements and/or scene changes on the images. And a breakdown of the map symbols and complex image may occur with camouflage. Map symbols are not easily recognized on complex image. Only with difficulty can you separate symbols from varying photographic representations of ground objects. Also when load map symbols onto complex background image, too much information may lead to symbol overload. The production of such maps is a big challenge, especially concerning the cartographic design. Therefore, it is very desirable to carry out a thorough investigation into the theories and key techniques for optimum cartographic design. From the problems occurred in the production of multi-scale image-map generation, the complexity-based strategies are proposed from four aspects: (a) the matching between the image and map, (b) the enhancement of background images, (c) the appropriate number of graphic symbols load on images, and (d) the appropriate label placement considering background images. To address the first factor, a complexity-based method for matching image resolution and map scale has been proposed for multi-scale image-map generation. The matching is based on the complexity of line features (line networks and individuals), as indicated by length, density, area, and fractal dimension. Experimental evaluations are conducted on 15 representative areas (urban, rural, and mixed) in Hong Kong at seven scales and eight image resolutions. Results show that the proposed complexity-based method can obtain good matching between image resolution and map scale in terms of accuracy and users' preference. To address the second factor, an experimental investigation has been conducted into the influence of the transparency of background images on the usability (i.e., effectiveness and efficiency) of image-maps. Image-maps with eleven levels of image transparency, at three scales and in nine areas have been designed and generated. An online questionnaire survey was conducted. A total of 1,263 participants took part in this experiment and they were asked to distinguish between natural and cultural features. Results show that (a) the usability of image-maps varies with the transparency of background images, mostly with a single peak; (b) the transparency level corresponding to the peak usability decreases as the background complexity increases. This serves as a guideline for the effective use of transparency in image-map design. To address the third factor, an experimental investigation into the effects of the complexity of background images and the density of graphic symbol on the usability of image-maps has been conducted to obtain the optimum level of map symbols load. Image-maps with ten density levels of map symbols from ten test areas in Hong Kong were used in eye tracking and online questionnaire survey. Forty participants took part in this experiment and they were asked to search Areas of Interest (AOIs) from image-maps/maps. Results show that (a) the optimum density level for may symbol load, where the peak usability occurs, varies with the complexity of the background images/image-map; (b) a polynomial model can fit the relationship between optimum level of symbol load and the complexity of the background image/image-maps. This can help optimize map symbol load for image-map generation. To address the fourth factor, complexity-based optimization is proposed for point label placement for image-maps. Both the general principles of point label placement and the complexity of background images were considered and integrated into the automated labeling optimization algorithm. The Boltzmann entropy of the patch (where the rectangular box of a label is located) is calculated as the complexity variable. Image-maps of three test areas in Hong Kong, which combined Points of Interest (POI) data and the corresponding satellite images, were used in the experiment. An experimental evaluation is conducted by a questionnaire in terms of "the ease level of finding the corresponding name labels", "congestion level" and "the satisfaction level". The proposed complexity-based strategy for point label placement has better performance in target searching tasks and satisfaction than original placement and strategy with general principles. In summary, this project has developed some complexity-based optimization methods of cartographic design for multi-scale image-map generation. In further research, more factors (e.g., automatic feature extraction methods, more image generalization methods, automatic symbol style design) can be explored. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/11809