This paper considers an approach for analyzing fibrillar collagen structures from electron microscopy images. It enables the quantitative comparison between collagen structural data (electron-optical data) and chemical data. The particular objectives of the paper are to model the electron microscopy images according to the periodic structure of collagen, provide methods for extracting periodic features directly from the experimental data and propose schemes for comparing these features with the theoretical amino-acid distributions of the examined collagen tissue. Theoretical models in the form of sequence-generated histograms are used as reference for extracting and analyzing the structural unit in images from collagen fibrils. In this respect, collagen provides a valuable model system for studying the chemical basis of ultra-structure and the mechanisms of various treatments on a protein, as well as detecting the alterations in collagen fibril structure produced by a disorder. The algorithms developed in this study can be applied to any fibrous protein, provided that its amino acid sequences and structural properties are known. Several application examples are presented. The algorithmic results are compared with clinical studies as to verify the applicability and potential of the proposed methodology.