The input natural skin images are used to predict the classification of skin diseases, and common diseases such as acne, actinic keratosis, alopecia androgenetica,alopecia areata, bullous dermatosis, chloasma, corn, dermatofibroma, eczema dermatitis, erysipelas, erythema multiforme, folliculitis, furuncle, haemangioma, herpes, herpes simplex, iga vasculitis, keloid, keratosis follicularism, lichen planus, lupus erythematosus, molluscum contagiosum, nevus, paronychia, pityriasis alba, pityriasis rosea, prurigo nodularis, psoriasis, rosacea, sebaceous cyst, sebaceousnevus, seborrheic dermatitis, seborrheic keratosis, skin tag, stasis dermatitis, cyringoma, tinea capitis, tinea corporis, tinea cruris, tinea manuum, tinea pedis, tinea unguium, tinea versicolor, urticaria, urticaria papular, varicella, verruca plana, verruca vulgaris and vitiligo can be detected.