The invention discloses a method for identifying counterfeit goods data based on an online dictionary learning data matching model, the technical proposal of the invention is that the hardware equipment is installed and debugged at first, networking and learning from images of counterfeit data matching models are performed, the data matching model of online dictionary learning is established, theversion of the goods to be inspected, Batch, packing and time patterns are placed beneath the scanning device; collection of image information and similarity matching are carried out through online dictionary learning data matching model, so that that final similarity is obtained, a threshold is set, and the final similarity is compared with the threshold. As the final similarity is greater than the threshold value, it is judged that the identify goods are authentic, when the final similarity is less than the threshold value, it is judged that the goods are counterfeit, and the information ofthe genuine goods and the counterfeit goods is saved for self-adjustment and self-proofreading, which effectively improves the accuracy of the identification judgment, saves the running time of the whole algorithm, and greatly improves the efficiency and success rate of the identification.