UAV Imaging and Object Based Image Analysis for early weed mapping
The UAV imageries are becoming very popular due to short time involved in the process (image capture and analysis) for decision making in agriculture. Recently, one very good research paper was published that shows how UAV image and Random forest based-object oriented image analysis would allow timely weed control during critical periods. This timely weed control is crucial for preventing yield loss.
In this context, a robust and innovative algorithm that involves object-based image analysis (OBIA), Digital Surface Models (DSMs), ortho-mosaics and machine learning techniques (Random Forest, RF), was developed by Castro et al., 2018, (MDPI, Remote Sensing).
Fig. Methodology adopted for automatic weed detection. (Credit :Castro et al, 2018)
Figures taken with thanks from the research work of Castro et al, 2018.
1- the OBIA algorithm identified every individual plant (including weeds and crop) in the image and accurately estimated its plant height feature based on the DSM.
2- the RF classifier randomly selected a class balanced training set in each image, without requiring an exhaustive training process, which allowed the optimum feature values for weed mapping to be chosen.
3- The system’s ability to detect the smallest weeds was impaired in images with lower spatial resolution, showing the necessity of 0.6 cm of pixel size for reliable detection, feasible in the UAV mission.
Ana I. de Castro, Jorge Torres-Sánchez , Jose M. Peña , Francisco M. Jiménez-Brenes , Ovidiu Csillik and Francisca López-Granados , 2018. An Automatic Random Forest-OBIA Algorithm for Early Weed Mapping between and within Crop Rows Using UAV Imagery. Remote Sensing 2018, 10(2), 285; doi:10.3390/rs10020285