For Raster Dataset, DCT encoding, SGL encoding or no encoding can be used for storage in SuperMap. Uncoded is the simplest, most intuitive and very important method of raster structure coding, which regards Raster Data as a data matrix and records pixel values row by row (or column by column). This storage method does not use any compression method, so when there are many grid pixels, it needs a considerable storage space, and with the improvement of spatial resolution, the storage space increases in geometric series. Because of the strong correlation in the spatial domain of the image, the adjacent pixels often have the same attribute value, and the use of certain compression coding can greatly save the storage space of Raster Data.

• DCT (Discrete Cosine Transform), Discrete Cosine Coding, is a widely used transform coding method in image compression, which provides a good balance between the compression ability of information, the quality of reconstructed image, the adaptation range and the complexity of algorithm. It has become the most widely used image compression technology. Its principle is to reduce the very strong correlation existing in the original spatial domain representation of the image through transformation, so that the signal can be expressed more compactly. The method has high compression ratio and performance, but the coding is distorted. Since the Image Dataset is not typically used for precise analysis, the DCT Encode Type is the compressed Encode Type that the Image Dataset stores.
• SGL (SuperMap Grid LZW), a compressed Storage Format customized by SuperMap. Its essence is an improved LZ W Encode Type. LZW is a widely used dictionary compression method, which was first used in the compression of text data. The encoding principle of LZW is to replace a string with a code, and the same code is used for subsequent identical strings. Therefore, the Encode Type can not only compress duplicate data, but also compress non-duplicate data. SGL improves LZW and is a more efficient way of compressed storage. At present, the compression storage of Raster Dataset in SuperMap adopts the compression Encode Type of SGL, which is a lossless compression.

When Grid Data is imported to Datasource as Dataset, and in 3D simulation, Generate DEM Data set is saved as Image Dataset in Current Windows. The compressed Storage Format of Raster Dataset can be selected and set; the Encode Type of Raster Dataset generated in Raster Analysis function is not Encode Type by default, and the setting is not supported.

The compression Encode Type corresponding to images of different types of Pixel Format in SuperMap is shown in the following table: