무인비행장치를 이용한 작물 생육 라이브러리 구축

Abstract

This study aimed to secure the video of crops cultivate in various times by using the Unmanned Aerial Vehicle to confirm the possibility of image reading for each crop, and structure the growth library for growth analysis. Structuring of data library is made by selecting 4 types of crops in the subject area to structure the orthophotograph as the base data for crop reading in accordance with the growth period and flight height. As a result of the visual reading by using the pattern of photo imaging for each elevation, the initial crop reading following the elevation and growth period had low recognition rate, but it gets heightened as it approaches to mid to later time of growth. In addition, the case of crop cultivation area of 1,000㎡ or more and the learning effect by pattern analysis have worked as an important factor in the increase of recognition rate. The result of this study is considered to be facilitated in the agriculture and statistics fields for the growth of crop and field survey works, and there is a need to reinforce the study on pattern recognition to overcome the limitation of visual reading structuring of various libraries for enhancing the crop recognition rate.