Photoelasticity images and videos contain information about the stress distribution in birefringent materials. Yet, its processing has difficulties in regions with high-stress concentrations. Making the material fail under load application. This paper identifies such regions by analyzing the intensity changes contained in photoelasticity videos. The evaluation is by the summa of each Euclidean distance between a frame, and the previous. The summa produces a grayscale image. Which is segmented using a thresholding process based on the range of intensity changes. It allowed separating regions with higher summa of color changes. That in this case, it corresponds to the zones with high-stress concentrations reported for a disc under compression. This technique allowed identifying high-stress concentration avoiding the sub-processes used in traditional photoelasticity.