Diabetic Foot Ulcers are an extreme compilation of diabetes, significantly impacting a patient's quality of life and leading to amputation and foot pathogenesis if not diagnosed and treated on time. Traditional clinical methods for Diabetic Foot Ulcer classification can be enhanced using deep learning techniques, yielding improved results; however, challenges include limited image data, artifacts, and high computational costs, especially in multi-class classification. Our research unfolds in two distinct phases. Initially, we collected datasets from various sources, while the subsequent phase delved into evaluating diverse Convolutional Neural Network algorithms for multi-class foot ulcer classification. This step performs preprocessing, such as enhancing ulcer region, artifact removal, correcting poor color illumination, ulcer area segmentation, suitable feature selection, and multi-classification of foot ulcers. Convolutional Neural Network techniques such as HYBRID CNN, HYBRID RCNN with YOLOv3, and YOLOv4 are used to achieve better accuracy. Furthermore, we attained the highest accuracy of 99.83\%, 98.06\%, and 97.88\% for Ischaemic, Neuropathic, and Neuro-Ischaemic ulcers, respectively, with YOLOv4. Finally, we summarized our research with an overview of the future trends and challenges in foot ulcer detection classification.