Automated pollen identification using microscopic imaging and texture analysis

Abstract

Pollen identification is required in different scenarios such as prevention of allergic reactions, climate analysis or apiculture. However, it is a time-consuming task since experts are required to recognize each pollen grain through the microscope. In this study, we performed an exhaustive assessment on the utility of texture analysis for automated characterisation of pollen samples. A database composed of 1800 brightfield microscopy images of pollen grains from 15 different taxa was used for this purpose. A pattern recognition-based methodology was adopted to perform pollen classification. Four different methods were evaluated for texture feature extraction from the pollen image: Haralick's gray-level co-occurrence matrices (GLCM), log-Gabor filters (LGF), local binary patterns (LBP) and discrete Tchebichef moments (DTM). Fisher's discriminant analysis and k-nearest neighbour were subsequently applied to perform dimensionality reduction and multivariate classification, respectively. Our results reveal that LGF and DTM, which are based on the spectral properties of the image, outperformed GLCM and LBP in the proposed classification problem. Furthermore, we found that the combination of all the texture features resulted in the highest performance, yielding an accuracy of 95%. Therefore, thorough texture characterisation could be considered in further implementations of automatic pollen recognition systems based on image processing techniques.

Forecasting pollination declines through DNA barcoding: the potential contributions of macroecological and macroevolutionary scales of inquiry

“…Previous attempts to tease actual pollination apart from observations of visits have revealed large differences in pollen transport networks vs visitation networks (Ballantyne et al, 2015), yet these studies have had to resort to painstaking visual pollen identification procedures that require rare alpha taxonomic expertise. Although there are automated visual approaches, they are also expensive, time consuming, and need just as extensive a reference database as does DNA barcoding (Marcos et al, 2015). Abundance and composition of heterospecific pollen transfer (HPT) has been determined to affect seed production and may underlie why oligolectic (specialist) pollinators heighten plant fitness more so than generalists (Arceo-G omez & Ashman, 2011; Brosi & Briggs, 2013).…”

Automatic and Online Pollen Monitoring

“…Some alternatives to manual quantification have been developed. Most methods are based on image recognition [18,19,20,21,22], whereas others used biomolecular analysis, i.e. DNA [23] or chemical identification [24].…”

Features Extraction for Pollen Recognition Using Gabor Filters

“…In both cases, the highest reported classification accuracy is about 87%. Marcos et al (2015) have achieved a better classification result -95%. In their study the authors have combined features derived from the following four features extraction methods: Haralick's grey-level co-occurrence matrices, log-Gabor filters, LBP and discrete Tchebichef moments.…”