To overcome the problems of low quality of image segmentation, as well as significant loss of the data, it seems promising to retain the data inaccuracies as realistic as possible during the image analysis procedures, instead of making hard decisions in the segmentation phase. Such an approach initializes the interest for new image analysis methods, handling grey-level images. Our work on developing shape analysis methods for fuzzy segmented images has resulted in a theoretical foundation for estimators of quantitative properties of digitized objects with fuzzy borders. In this paper, we present results of perimeter, area, and compactness measure estimations obtained by applying a fuzzy approach to digitized objects in low resolution real images.