1) The document proposes a methodology to recognize facial emotions from images using edge detection and comparison of feature regions between normal and emotional expressions. 2) The key steps involve preprocessing images, cropping regions of interest around the eyes, eyebrows and lips, applying edge detection, and comparing deviations in edge positions between expressions to classify emotions. 3) Experimental results showed 83.33% accuracy in classifying happy, angry and null expressions, with higher accuracy for happy and null emotions compared to angry emotions.