3D data optimization

Optimization methods

SuperMap 3D takes a series of measures in order to optimize 3D data and improve the data transmission speed and service access efficiency. The optimization methods are different for different types of data, and are introduced detailedly here.

1. Vector data

For the transmitted vector data, SuperMap 3D adopts following methods to improve the data transmission efficiency: storing 2D points, but not 3D points.

A third-dimensional coordinate (elevation z) is added in the process of converting the 2D vector data to the 3D data, and thus the data quantity is notably increased. Therefore, directly transmitting 2D vector data can effectively reduce the data quantity without affecting the quality of the vector data.

In addition, in order to improve the precision of the storage nodes, SuperMap 3D adopts the method of changing the data types of vector data coordinates for the transmitted vector data. That is, the short (8 bits) integer which is the original format of vector data coordinates is changed as float (32 bits).

2. Image data

For the tile data format in the image cache, SuperMap 3D considers the advantages and disadvantages of the JPEG and PNG file formats and thus provides the image cache methods of mixed formats, in order to improve the data transmission speed and reduce the data transmission quantity. Here, the JPEG image format (*.JPG, *.JPEG) is a common format of lossy compression, which has powerful compression functionality and can compress the image into a small storage space. The PNG (*.png) format is the most popular file format in the WEB applications, which supports high-level lossless compression and provides lossless-compression image files that are 30 percent shorter than those of the GIF format. Compared with the lossy compression of JPEG, the compression of PNG is less in amount but can keep edge transparent or semi-transparent.

Based on above features, SuperMap 3D provides the cache of JPEG+PNG mixed mode, which permits users to use the tiles of both JPEG and PNG image formats in the same cache. Users can choose to use JPEG format for the tiles in the centre of the cache area so as to reduce the relevant file size, and use PNG format in the periphery of the cache (i.e. this part of tiles need to keep transparent). This mixed mode considers the advantages of both JPEG and PNG image formats. The cache files maintain the JPEG's advantage of occupying small disk space and make the whole image display not affected by the edge of the cache tiles, thereby effectively reducing the disk space occupied by the cache files and improving the image data transmission speed.

3. Terrain data

For the transmitted terrain data, SuperMap 3D adopts following methods to reduce the data transmission quantity and improve the data transmission efficiency:

• When partitioning the terrain data, the sample points increase a lot, so each blocked data has a large data information. By integrating the blocked data into a large file, the quantity of blocked data, files in the disk reduce a lot. Thus data transmission times and cache creating time are shortened to improve the browsing efficiency on the client side;
• Compressed transmission. That is, the data is compressed by the server before transmission. The client should decompress the downloaded data before using it.

4. Model data

SuperMap 3D adopts the direct compression method to improve the model data transmission efficiency. That is, the server generates the data by compression mode and transmits the compressed model data to the client, and then the client decompresses the downloaded data before using it. Such a process of compression and decompression can reduce the data transmission quantity and thus improve the transmission performance.

Comparison of the performances before and after the optimization

Comparison of the download time and rendered frame number before and after the SuperMap 3D client optimizing the downloaded image data, model data, terrain data and vector data.

As shown above:

• The compression scale of terrain data is bigger, followed by that of model data and image data. The compression of vector data has few effects.
• For terrain data, the downloaded data and download time change relatively notably before and after the compression.