Abstract: Skin cancer, particularly Melanoma, poses a significant global health threat. Early detection is crucial for reducing its impact. Existing research on melanoma detection lacks speed and accuracy. Our proposed system combines DEECO for contrast enhancement and DMFFX for classification. We address drawbacks like over-fitting with multi-level feature fused Xception. Segmentation utilizes AAMBCS for precise region extraction. Using ISIC 2016 dataset, our model's efficacy is measured against conventional methods. Intended to aid oncologists and dermatologists, our model aims to improve skin cancer treatment outcomes.