A new colour vision system to quantify automatically foliar discolouration caused by insect pests feeding on leaf cells. (May 2015)

• Record Type: Journal Article
• Authors: Clément, A.; Verfaille, T.; Lormel, C.; Jaloux, B.
• Is Part Of: Biosystems engineering. Volume 133(2015:May)
• Page Start: 128
• Page End: 140
• Publication Date: 2015-05
• Subjects: Colour imaging -- Histogram analysis -- Damage classification -- Plane tree -- Sycamore lace bug; Bioengineering -- Periodicals; Agricultural engineering -- Periodicals; Biological systems -- Periodicals; Génie rural -- Périodiques; Systèmes biologiques -- Périodiques; 631
• DOI: 10.1016/j.biosystemseng.2015.03.007
• Languages: English
• ISSNs: 1537-5110

Abstract

Abstract : Leaf colour and leaf area are key parameters for many plant studies, including plant protection. The quantification of foliar bleached area is often used to evaluate the strength of plant attack by pests or diseases, to compare the efficiency of different treatments, the resistance of some populations, and to set up some control methods. The quantification of such area by human vision often lacks accuracy and reliability. Image analysis systems could bring a more accurate and objective measure, and could be automated to treat a great number of samples. Such an automated tool has been developed in order to measure quantitatively foliar bleaching due to the sycamore lace bug, Corythucha ciliata (SAY) on Plane tree. This tool was built up with a colour image capture bench and a fast automated segmentation process based on an original chlorophyll histogram unsupervised classification. Dedicated software was developed and integrated in an operational image processing platform capable of routine use by non-specialists in image analysis. The accuracy of the tool was determined by comparison to human expert segmentation. A very low error rate was observed in the absence of artefacts, but artefacts such as powdery mildew symptoms were not well distinguished and lead to weaker performance. Comparing its reliability and robustness to classical visual estimation and classification method, the tool performance was similar to the most experienced rater. The advantage of such a system is the possibility to treat automatically a large number of pictures and produce accurate, reliable, repeatable and non-subjective quantitative measurements. Highlights: We developed an innovative image analysis tool to quantify foliar damages. The automated process enables to treat quickly a huge number of samples. Its accuracy was validated comparing its performance to expert segmentations. It is more reliable than the classical visual estimation and classification method. The tool is potentially transposable to other kinds of pests or foliar diseases.