Non-invasive assessment of bone strength is an important component of biomechanics research. Soft tissue structures like trabeculae in femur bones are to be assessed precisely for tissue modeling and fracture mechanics. The strength and architecture of these structures in femur bones are routinely analyzed by varied image based methods for diagnosis and monitoring of osteoporosis like metabolic disorders. In this work, an attempt has been made towards development of an automated system for analysis of trabecular femur bone architecture using spatial frequency decomposition method. Conventional Radiographic femur images recorded using standard protocols are used for the study. The compressive and tensile regions in the images are delineated using pre-processing procedures. The delineated images are decomposed into their corresponding Intrinsic Mode Functions using bi-dimensional empirical mode decomposition method. The characteristic feature vectors corresponding to regional variations are used for further analysis. The result shows that the extracted features are distinct for compressive and tensile regions and are found to be consistent for the first three modes. The values are high for compressive regions. The higher modes are not observed in tensile regions of abnormal images which are attributed to the loss of structural patterns. As the strength of the bone depend on architectural variation in addition to bone mass, this study seems to be clinically useful. The objectives, methodology and analysis are presented in this paper.