QUANTITATIVE ASSESSMENT OF LEAF DAMAGE IN AESCULUS HIPPOCASTANUM: A COMPARISON OF VISUAL AND SOFTWARE-BASED METHODS FOR EVALUATING CAMERARIA OHRIDELLA INFESTATION

Abstract

The invasive moth Cameraria ohridella significantly threatens Aesculus hippocastanum (horse chestnut) trees, causing extensive leaf damage that undermines urban aesthetics and tree health. Traditional visual evaluations of leaf damage are time-intensive and prone to subjectivity, resulting in variable accuracy. The aim of this study was (i) to test whether the accuracy of visual damage assessment depends on the level of biological expertise of evaluators, and (ii) to determine whether a software-based method provides faster and more reliable estimates than human assessment. This study compared visual assessments by laypeople and experts to a novel software-based method for evaluating leaf damage. Leaves were collected from multiple sites, scanned, and analyzed manually. In addition, a software algorithm based on digital image analysis was developed, which automatically segments leaf images and quantifies damaged areas using color-based pixel classification. Statistical analysis revealed significant differences between laypeople and expert assessments (p < 0.05) and between laypeople and the software (p < 0.01). No significant differences were found between software and expert evaluations (p > 0.05). The software method reduced evaluation time by 97.51%, with manual analysis of 105 leaves averaging 40.22 minutes per person, compared to under 3 minutes using the software. The software-based method offers a reliable, efficient, and scalable alternative to traditional visual assessments, particularly for large-scale monitoring. However, the study is limited to scanned leaf images collected under controlled conditions, and further validation under variable field conditions is needed. By providing faster and more consistent evaluations, it supports improved management of Cameraria ohridella infestations, aiding urban forestry efforts to preserve the health and aesthetics of horse chestnut trees.