Diabetic Foot Ulcer (DFU) is characterized by deep tissue lesions linked to peripheral vascular disease and neurological conditions in the lower boundaries. According to National Institute for Health and Clinical Excellence methods early and efficient management of DFU is essential to improve overall quality of life. To classify the DFU image and healthy foot skin the proposed Focal Attention MobileNet (FAM) is developed. Anisotropic Diffusion Filter (ADF) is applied to the DFU dataset that remove noise in the DFU image edges. Using Fuzzy Adaptive Thresholding (FAT) the image region’s followed by adjusting the threshold values. Cognitive Contour Descriptor (CCD), extract features by shape, area, boundary box, perimeter, centroid of DFU for further analysis. Finally, FAM is proposed that enhance the classification of DFU images, in foot analysis. Using Python software, the DFU image randomly chosen from DFU dataset and the proposed FAM achieved a higher accuracy of 96% and similarly scores of precision, recall achieved about 96% which is highly efficient compared to the existing method.