With the development of the 4th industrial revolution, much research is underway to introduce the latest ICT technology in agriculture. In the case of the bare ground agriculture, however, it is difficult to control the environment in order to work from the outside, and diseases caused by pests are the most problematic. Currently, IT pheromone traps are being used by farmers to solve the above problems, but it takes a very long time to send the results to the experts to receive the diagnosis and the result of the action, which is problematic because of the high labor consumption. In order to solve the above problem, we try to understand the number of insect pests by analyzing the image through the proposed CNN method by taking pest image using Raspberry pie with low installation cost. For CNN analysis, pest images were extracted from national organizations and Google images, and images were extracted from farms where actual pests were generated. The insufficient pest image is a system that rotates the image in eight directions, first extracts the characteristics of the pest from raspberry pi, and secondly analyzes the image through the cloud server. By using the above method, it is expected that it will reduce productivity of pests of farmers and increase the productivity and reduce the labor force of advanced manpower of professional organizations.