There are several way in the literature. The classical way is the analysis of fringer order (Caputo and Standlee 1987). Another way is to use a specific machine (Kratos Ind., Sao Paulo, Brazil) to calculate the stress in a specific point, that according to Dally and Riley 1978, the “stress optical law” for the plane photoelastic model is used to determine the individual maximum shear stress (τ) of each point. It is represented by the following equation: τ= (kσN)/2h (Mpa).
First of all: You have to see what kind of image you have taken (isochromatic or isoclinic) [To know the details refer the monograph "Experimental stress analysis" by Dally and Riley]
If the photoelastic image is an isochromatic, it has the information about the difference of principal stresses only, which can be determined by measuring the fringe order using a proper compensation technique (refer to the book above again for details). If you need individual components of principal stress; you need to determine both the isochromatic and isoclinic. If you can share your photograph, I would be able to help you.
what you need to get the principal stress difference is the phase retardation which can be retrieved from the birefringent extinction you measure in a cross polarized image. I suggest finding a book and carefully go through the concepts. My favorite is "photoelastic stress analysis" by A. Kuske (ch. 5&6).
There are many standard text books on Photoelastic Stress Analysis. The basic principles involved and techniques followed for stress analysis are well explained.