Water stress detection based on optical multisensor fusion with a least squares support vector machine classifier Water stress detection based on optical multisensor fusion with a least squares support vector machine classifier

The objective was to optically discriminate between healthy and water stressed wheat canopies. Canopies were grown under greenhouse conditions. The aim was to develop an optical multisensor system that can detect and identify biotic and abiotic stresses. In the current investigation the successful recognition of water stressed and healthy winter wheat plants in the presence of a Septoria tritici infection is presented. The difference in spectral reflectance and fluorescence response between healthy and stressed wheat plants was investigated. Stress type detection algorithms have been developed based on the combination of least squares support vectors machine (LSSVM) with sensor fusion. Through the use of LSSVM, classification performance increased to more than 99%. These results show promise for the development of cost-effective detectors for automated recognition of different biotic and abiotic stresses.