Digitizing from a aerial photo is commonly referred to "heads up" digitizing.
If the image is already georeferenced, then you can use many GIS software packages to create new layers from the image. You will need to learn from your software how it handles overlapping lines and polygons in order to make sure you use good methods when digitizing. You will need to know how the image was acquired and georeferenced to make sure that you understand where some of the sources of potential error can occur.
If the image is not georeferenced, then you must first georeference it before being able to digitize a feature on it. There are many methods to do this, and most involve setting up known coordinates on the photo either from ground control or from another already georeferenced data set. This process can be tricky and requires that you know something about the photo and the method of its acquisition. Photo images can have distortion on the edges due to lens errors and plane-based pitch and yawl. The methods for georeferencing imagery are more complex, and you should seek more information specific to rectifying or georeferencing aerial photography.
