The first step is to 'box out' the individual particles. That is, select the individual particles from the digitized micrograph (or CCD frame), and save each as a separate image. Currently EMAN doesn't have an internal program for doing this, but numerous programs are available for this purpose. For manual boxing of particles EMTOOL is suggested (it has many other uses as well). For automatic boxing, PTOOL is good (for certain classes of particles). Save all of the boxed particles in each micrograph (or frame) to a separate file in the IMAGIC format. That is, you should have one IMAGIC file for each micrograph/frame. This file (actually a pair of files, one '.hed' file and one '.img' file) contains all of the particle images from that micrograph.
Please note, when scanning images, it is important to use the proper resolution. For example, if you're trying to get a 15A structure, you should try to scan so the images have a pixel size of about 5A, that is ~1/3 of the final resolution goal. Please also keep the memory requirements of the software in mind. If your boxed particles are 200x200 pixels, the final 3d model will be 200x200x200x4 = 32 megabytes. Your computer must have enough memory to hold at LEAST 2 of these models in memory at once, preferably 3.
Next, select the type of particle you have data for:
Particle with 1 axis of rotational symmetry (ie - c3 or c4)
If the symmetry of the particle is unknown, you can use 'symhist' to try to get some idea of what, if any symmetry your particle might have.