An Automatic Building Label Placement Method For 3D Visualizations

Map label,as one of the basic contents of the map,can help users quickly understand the information in the map and enhance the readability of the features.Compared to a two-dimensional(2D)map,a three-dimensional(3D)scene has far greater detail and realism.However,these enhanced visualizations can introduce unnecessary distractions,making it difficult for users to recognize spatial features.Therefore,supplementary means are needed to enhance scene identifiability.For example,map symbols and text annotations are used to explain scene elements.Building objects are typical features in 3D scene,adding suitable building labels can greatly improve the recognizability of buildings and the readability of 3D city scenes.The annotation effect of 3D buildings is influenced by the attributes of the 3D label,building visibility,and the spatial relationship between the building and viewpoint.In this context,the optimal label position and orientation of each building can be calculated based on rational rules and formulas at runtime.However,such an optimal calculation method has a substantial complexity and a strong influence on performance.By contrast,evaluating candidate labels and then selecting the suitable label for each building can be effectively implemented.In order to enhance the readability of 3D building and improve the efficiency of label placement,this study introduces an approach for labeling buildings in 3D scenes based on evaluations of label candidates,performing view-dependent label evaluation,selection and placement during interactions.Based on this,we study the solution for resolving 3D building label conflict.The main research contents of the thesis are as follows:1)The study explores the characteristics of fixed-orientation labels and dynamic-orientation labels,constructs a candidate set containing a limited number of labels for each building.The fixed-orientation label can reflect the building's surface orientation,while the dynamic-orientation label can maintain high readability from varied viewing angles.According to the characteristics of the labels,the evaluating factors of the candidates are proposed.The labels are evaluated in real time according to the orientation factor,the word length factor,and the visibility factor.After that,the valid candidates of each building are retained.2)Considering local rank in the label set,label depth,building salience,and conflict degree,we study the label priority applicable to building features in 3D scenes.Based on this,a greedy algorithm is proposed to resolve label conflicts in screen space,achieving a trade-off between interaction efficiency and label layout quality.3)Based on the proposed method,the building label placement experiment was carried out to verify its practicality and effectiveness.In addition,we analyze the performance of the label method in detail,and compare it with existing methods.The research results show that the proposed method can adapt to the dynamic changes of perspective and distance of sight,and effectively enhance the correlation between labels and the annotated buildings.In addition,with the help of the greedy algorithm,which sequentially places label in order of priority,the label conflicts in screen space are efficiently resolved.The method improves the overall visualization readability by retaining labels with low depth,less conflict degrees,and higher building salience.The proposed 3D building label placement method provides a new technical approach for annotation in 3D urban scenes,and expands the research and application fields of 3D cartography and virtual geographic environments.