Processo de captura de imagens com VANT e segmentação de imagens na identificação de falhas na cultura da soja utilizando o algoritmo ckMeans

Abstract

The development of technologies for agriculture has made possible better control and management production systems. Through monitoring, it becomes possible to analyze phytosanitary conditions, as well as productive potential, and especially failures resulting from process in crop implementation and its management. Thus, this present work aims to identify plant stand failures in a soybean crop grown in floodplain areas with image capture and processing. These were captured with aid an unmanned aerial vehicle and processed by the ckMeans algorithm, reducing number of colors in relation to the original image. Thus, with processing it is possible to identify crop line failures, management errors, sowing, germination problems and areas with drainage deficiency. Therefore, using processed images is a tool that contributes to decision making in soybe crop anmanagement cultivated in lowland areas.