The work aims to solve the problem that traditional brightness, color and other detection methods can not effectively detect whether the cigarette pack (especially the white background of the cigarette pack) is oil-sealed.On the basis of traditional detection methods, the detection method based on gray scale co-occurrence matrix algorithm was added for bulk pack detection. Through the novel hardware architecture design of bulk packet detection, hardware isolation design, camera optimization, mathematical modeling, eigenvalue extraction of gray scale co-occurrence matrix, optimal solution and other processing, the novel bulk pack detection method based on gray scale co-occurrence matrix algorithm could improve the detection accuracy of cigarette pack and effectively detect whether the cigarette pack (especially the light-colored cigarette pack) was oil-sealed, and reduce the flow of unsealed packs into the next process.The novel detection method based on gray scale co-occurrence matrix algorithm could effectively improve the detection accuracy of cigarette pack, and the detection and rejection rate of cigarette pack was ≥99.9%. The novel detection method based on gray scale co-occurrence matrix algorithm can effectively improve the detection accuracy of cigarette pack, especially for light-colored cigarette pack with or without oil seal.