You have several options. The easiest is to photograph the microscopic image, place a transparent grid (physically or on the computer screen). a grid with 100 cells usually is sufficient. Count the number of algae under a grid cell intersection, where the lines cross. Then count the points where the anemone tissue or empty space occurs. The total is 100. You can report either as a percentage of area or calculate the area in the slide and covert to mm of area occupied by algae.
This can also be done with a grid that sits in the ocular of the microscope, that is the old fashioned way before there were computers and film was expensive.
Other methods include sacrificing the anemone, isolating the algal cells and counting as you would a hematocrit, which uses a special slide (hemocytometer) that also has grid lines.
Counting a sub-sample using a hemocytometer works well. Also, when you isolate your zooxanthellae through homogenization and centrifugation, keep the supernatant... you can measure the protein in that and standardize number of zoox to mg of protein.
Alternatively, depending on the size of your anemone, you can use in vivo fluorescence to estimate zoox numbers for the entire animal. The technique was used by Burner et. al. 1993 in their study of repopulating an aposymbiotic anemone (J. Exp. Mar. Biol. Ecol., 170: 145-158.