Using image technology to recognize the stored-grain insects, the classification and recognition of multi-characteristic parameters and multi-compound degree of various pests must be settled. This paper presents a new approach—extension engineering method—to solve this problem. This approach is based on constructing standard and extensional matter-element matrixes for various pests classification using matter-element and extension set, calculating the correlative degrees between the stored-grain pests to be recognized and various pests, and recognizing the pests according to magnitude of correlative degree. According to the calculated result, it is proved to be an effective and practical approach to classification and recognition of stored-grain pests.