Pollen Classification Based on Geometrical, Descriptors and Colour Features Using Decorrelation Stretching Method

Abstract

Abstract. Saving earth's biodiversity for future generations is an important global task, where automatic recognition of pollen species by means of computer vision represents a highly prioritized issue. This work focuses on analysis and classification stages. A combination of geometrical measures, Fourier descriptors of morphological details using Discrete Cosine Transform (DCT) in order to select their most significant values, and colour information over decorrelated stretched images are proposed as pollen grains discriminative features. A Multi-Layer neural network was used as classifier applying scores fusion techniques. 17 tropical honey plant species have been classified achieving a mean of 96.49% ± 1.16 of success.

Pollen segmentation and feature evaluation for automatic classification in bright-field microscopy

“…Without a doubt one of the major efforts in this kind of studies has to do with the compilation, preparation and labeling of datasets. Thus, some of those studies deal with 300-500 total samples and/or 3-5 pollen types (Rodriguez-Damian et al, 2006;Chen et al, 2006;Boucher et al, 2004;Carrión et al, 2002;Travieso et al, 2011;Ticay-Rivas et al, 2011). The study from Chica (2012) is one the most complete in this respect with 5 pollen types and 1063 total pollen grains.…”

“…A previous work in the area of aeropalynology (ASTHMA EU project) used multifocus stacks and reported recognition rates around 97% for 5 pollen types (Boucher et al, 2004). Other studies demonstrate accuracy ratios between 90% and 97% (Rodriguez-Damian et al, 2006;Chen et al, 2006;Ticay-Rivas et al, 2011;Chica, 2012;. However, such ratios must be considered with care, they are not reliably comparable because their training database usually differ largely in terms of pollen genre and/or number of training samples, which is directly related to obstacle (b).…”

Principles and methods for automated palynology

“…Highest success rates have generally been achieved with a combination of image processing and a machine learning‐based (neural network) classifier (i.e. Li et al ., ; Ticay‐Rivas et al ., ). However, the highest number of taxa classified so far was achieved by Ronneberger et al .…”

“…More recently, Ticay‐Rivas et al . () have achieved even better success rates (>94%) in a 17‐type pollen set.…”

“…Highest success rates have generally been achieved with a combination of image processing and a machine learning-based (neural network) classifier (i.e. Li et al, 2004;Ticay-Rivas et al, 2011). However, the highest number of taxa classified so far was achieved by Ronneberger et al (2002), using processed images of fresh pollen collected through a combination of fluorescence and confocal microscopy.…”