In April 2016 Manchester eScholar was replaced by the University of Manchester’s new Research Information Management System, Pure. In the autumn the University’s research outputs will be available to search and browse via a new Research Portal. Until then the University’s full publication record can be accessed via a temporary portal and the old eScholar content is available to search and browse via this archive.

Related resources

Search for item elsewhere

University researcher(s)

Bruce Grieve's research staff profile

Academic department(s)

Localized Multispectral Crop Imaging Sensors Engineering & Validation of a Cost Effective Plant Stress and Disease Sensor

Grieve B D, Hammersley S, Mahlein A K, Oerke E C, Goldbach H

In: Gurkan, Deniz ; Baglio, Salvatore. IEEE 2015 Sensors Applications Symposium: IEEE 2015 Sensors Applications Symposium; 11 Apr 2015-15 Apr 2015; Bože Peričića 14, 23000 Zadar, Croatia. Piscataway, New Jersey, USA: IEEE Instrument & Measurement Society; 2015.

Access to files

FULL-TEXT.PDF (pdf)

Abstract

Close proximity hyperspectral and multispectral imaging of crops and soils offers significant potential to optimize sustainable intensification of arable produce and seeds breeding, through the real-time precision management of plant pathogens, viruses and pests and the non-destructive high throughput screening for beneficial crop traits. These opportunities have been recently reported and are the subject of ongoing R&D within industry and academia. The broad uptake of the technology by large commercial end-users, through integration with in-field and glasshouse machinery, is limited by cost and equipment reliability. It is further restricted by spectral and spatial resolution, power budget and size, when extending its applicability to consumer markets and small-holder farmers. This study verifies, for the first time, that multispectral sensor systems architectures, exploiting proprietary narrowband LEDs and silicon C-MOS imaging detectors, are capable of substituting for conventional and more expensive line-scanning hyperspectral imaging systems when operated in close proximity (c. 1-2m) of a crop canopy. This was achieved by comparing the data from a prototype version of the new LED-sensor system versus a reference laboratory hyperspectral imaging unit, which was previously developed for crop phenotyping, and the early detection of two fungal pathogen borne diseases in whole barley and sugar beet plants. The choice of crops and diseases replicates earlier studies, with the reference hyperspectral unit, and serves to demonstrate the generic applicability of the new LED-sensor system to cereal and tuber classes of crops. The results indicate that the new approach can deliver data of comparable quality to that of the reference system, for in-field duties, and offers the opportunity for higher sensitivity and spatial resolution. Future potential to apply the new multispectral, LED-based system within commercial products is then discussed.