Relationship between open source class library gdal and ArcGIS

ArcGIS 10 has come out, and rich GIS tools and improved interface interaction are new features of 10. These new features also support new data formats. The format is divided into two parts: grid data format and vector data format.

The vector library of ESRI supports the most powerful ArcGIS desktop platform in the industry. No matter from the abstraction of vector models to the standardized definition of data formats, ESRI is at the forefront of the industry. With the support of FME extension, almost all vector data formats are currently supported, and nearly 100 + vector formats are supported.

The processing of raster data ESRI is not as good as that of mainstream remote sensing software, but now it is very popular to catch up with the trend, such as algebraic operations in raster data, as well as limited classification tools, ArcGIS gradually shows comprehensive support for raster data. 10, imageserver has become a separate extension module, rather than a separate product, and the entire function has a better close relationship with ArcGIS.

If so much is said, what is the relationship with gdal open source. I would like to say that it is highly relevant. The new version of ArcGIS 10 supports more raster data formats, essentially because the new version of gdal raster Library supports more data formats. In fact, ArcGIS 10 uses the gdal class library to process raster data at the underlying layer.

What are the characteristics of gdal classes? Why does ESRI use it?

I think a good model abstraction is the key to the success of this open-source class library. Whether you are a grid or a vector, a good mode is the key to success. If you have read the design mode, you can understand the operation methods of the same process. All raster data operations in the raster also have a fixed operation method (in fact, this depends on the structure), because the essential operation of the raster definition is to modify the cell, and the details are slightly different, some are 1-bit, some are multiple-bit, some are all information in one band, and then the second band is saved. Some are information in one band followed by another band, then the second row of the first band and the second row of the second band are stored. All these are details that can be masked. In terms of representation, our understanding of raster data is simple. There is only one cell in one layer of data in one band. Each type of gdal raster data format is processed in this way. users do not need to understand the details, but only need to know the band (layer data), such ), perform operations on each cell of each layer of data. The cell value can be the actual color or elevation. This is especially true for vector abstraction. In ArcGIS, the vector model description design is layer, which contains datasets of multiple layers ), what we see in the vector library of gdal is also these concepts. The most fine-grained user only needs to know the feature. As for how to store the actual data under the feature, the user does not need the relationship. With these concepts, all different data formats can be treated in the same way. This is the power of model abstraction. Logical conceptual Division (called primitive operations in design patterns) abstracts the essential features of the general meaning of objects, the same operation can process all such different objects.

What can we do to understand the model abstraction of gdal?

Why do I study it? I want to extend it. The essence of expansion is to define the format defined by myself according to its model specifications.

What can I do if I am familiar with gdal class libraries?

You can write a GIS system by yourself, and it is very powerful. Because the essence of GIS is data processing, gdal is the kernel. Our work is limited to how to read and write data and how to display data. This makes it possible for individuals to write a GIS processing platform.