Geodeziya, kartografiya, geografiya

What do I need to create a land cover classification map?
There are several options for creating land cover maps. The following is a list of a few data and interpretation alternatives that can be considered for mapping land cover:
• Visual satellite image interpretation
• Digital satellite image interpretation
• Aerial photo interpretation using stereo photos
• Aerial photo interpretation using single uncorrected photos
• Aerial photo interpretation using orthophotos
• Field surveys using simple angle and distance measurements
• Interpretation of videography
• Interpretation of small format photography
This guide focuses on approaches using remotely sensed data but you can apply many of the steps to other data sets.
First off, creating a land cover map using remotely sensed imagery necessitates suitable imagery. There are several choices available and the process of selecting the appropriate imagery is described in the section "How do I select the imagery I need?"
You will also need some way to visualize and process the imagery. If the imagery is in a printed form then viewing it is relatively straightforward and the classification process is limited to visual interpretation methods. However, if the imagery is in digital form you will need software to view and process the imagery. Software required for classifying imagery can range in price from free to tens of thousands of dollars. There are several options available when using digital imagery and these are detailed in the section below titled "Which classification method should I use."
You will also need a certain skill level to create an accurate map. Arguably the most important skill is the ability to associate the features that can be seen in an image with what is on the ground. This ability comes from experience. A good way to begin learning this skill is to look at imagery from an area that you are familiar with and begin identifying features on the image based on your recollection of what you know exists on the ground. The importance of this type of experience cannot be over stated.
In the classroom emphasis is usually put on the mechanics of automated classification. This is certainly important information for a remote sensing practitioner, but understanding the details of how an algorithm works or the physics of remote sensing is not required to produce an accurate land cover map. There are some simple methods that work very well as long as the analyst has the ability to reliably identify features in the image.
Lastly, whether you are just learning image classification or you have been through some formal training such as workshops or university classes, it is always a good idea to get some feedback from experienced colleagues. Discussing your plan of action and periodically showing your progress to other remote sensing practitioners is very helpful. They can sometimes suggest refinements and offer other options that will help improve the land cover map.