The invention discloses a double camera-based intelligent human face identification method and system. The method comprises the steps of collecting a black-white human face image and a color human face image through double cameras respectively; performing saliency detection on the black-white human face image by adopting a frequency domain residual error method to obtain a black-white saliency image; performing black-white human face identification in the black-white saliency image by adopting a differential image-based PCA method; performing skin color extraction and segmentation on the color human face image by adopting an improved skin color detection algorithm to obtain a color human face candidate region; performing color human face identification in the color human face candidate region by adopting an improved 8-level parallel PCA method; and performing weighted identification according to a black-white human face identification result and a color human face identification result to obtain a final human face identification result. The method and the system have the advantages of high identification rate, accuracy and high robustness, and can be widely applied to the field of bio-feature identification.