The present application is applicable to the technical field of image recognition, and provides a reproduced image recognition method and system and a terminal device. The method comprises: constructing a reproduced image classifier according to a plurality of training samples, wherein the training samples comprise training images and corresponding classification results, and the classification results indicate real images or reproduced images; extracting feature values of an image to be classified, wherein the feature values comprise a Y-channel brightness conversion rate of said image and a surface gradient feature value of said image; and carrying out, by means of the reproduced image classifier, classification discrimination on the feature values of said image, and recognizing, based on a classification result, whether said image is a reproduced image . The trained reproduced image classifier is used to recognize, according to the Y-channel brightness conversion rate and the surface gradient feature value of the image, whether the image is the reproduced image, which makes it possible to effectively and intelligently discriminate whether the image is the real image or the reproduced image, thereby effectively avoiding faking behaviors, and solving the problem of being unable to accurately distinguish the reproduced image and an actually photographed image at present.
展开▼
