New strategies for row-crop management based on cost-effective remote sensors

Abstract

Agricultural technology can be an excellent antidote to resource scarcity. Its growth has led to the extensive study of spatial and temporal in-field variability. The challenge of accurate management has been addressed in recent years through the use of accurate high-cost measurement instruments by researchers. However, low rates of technological adoption by farmers motivate the development of alternative technologies based on affordable sensors, in order to improve the sustainability of agricultural biosystems. This doctoral thesis has as main objective the development and evaluation of systems based on affordable sensors, in order to address two of the main aspects affecting the producers: the need of an accurate plant water status characterization to perform a proper irrigation management and the precise weed control. To address the first objective, two data acquisition methodologies based on aerial platforms have been developed, seeking to compare the use of infrared thermometry and thermal imaging to determine the water status of two most relevant row-crops in the region, sugar beet and super high-density olive orchards. From the data obtained, the use of an airborne low-cost infrared sensor to determine the canopy temperature has been validated. Also the reliability of sugar beet canopy temperature as an indicator its of water status has been confirmed. The empirical development of the Crop Water Stress Index (CWSI) has also been carried out from aerial thermal imaging combined with infrared temperature sensors and ground measurements of factors such as water potential or stomatal conductance, validating its usefulness as an indicator of water status in super high-density olive orchards. To contribute to the development of precise weed control systems, a system for detecting tomato plants and measuring the space between them has been developed, aiming to perform intra-row treatments in a localized and precise way. To this end, low cost optical sensors have been used and compared with a commercial LiDAR laser scanner. Correct detection results close to 95% show that the implementation of these sensors can lead to promising advances in the automation of weed control. The micro-level field data collected from the evaluated affordable sensors can help farmers to target operations precisely before plant stress sets in or weeds infestation occurs, paving the path to increase the adoption of Precision Agriculture techniques.