Over the last decade, lotus root plants and rice paddies have been seriously damaged by the golden apple snail (GAS) as an invasive species. They feed on young lotus root and rice seedlings. The number of GAS that succeed in overwintering is increasing rapidly, and exterminating GAS egg masses is becoming increasingly important. However, the approach required to exterminate the egg mass differs depending on their state in the incubation process. To achieve automated extermination of such egg masses, a method for automatically identifying the state of the incubation process of an egg mass is required. Therefore, in this study, we present an image-based classification method that considers the incubation process of egg masses. In the proposed method, semantic segmentation is first applied to detect only the egg mass. Next, the color information of the detected egg mass is analyzed, and the state of the incubation process of the egg mass is classified. We conducted an experiment in which we verified the effectiveness of the proposed method for images of egg masses from an agricultural canal and evaluated its detection and classification accuracy.