Discernment of bee pollen loads using computer vision and one-class classification techniques

Chica, Manuel; Campoy, Pascual

doi:10.1016/j.jfoodeng.2012.03.028

Cited by 19 publications

(15 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The pollen images of C . pachystachya had 31 hits (85.57%) with this technique, a percentage that is very close (89%) to that obtained by [ 4 ] in a study on Urticaceae pollen grains.…”

Section: Resultssupporting

confidence: 85%

“…We obtained a performance of 64% with the CST+BOW technique. In spite of the general low performance compared to studies that utilized less pollen types [ 2 , 4 , 6 , 8 ], seven pollen types that we classified with the CST+BOW technique had performance ≥80%, A . colubrina and S .…”

Section: Discussionmentioning

confidence: 71%

“…In researches aiming to automate the classification of pollen grains, the performances reported by [ 2 ] with three species achieved 90% with the technique of neural network. [ 4 ] have used four pollen types and reached 94% with the single classification class technique. [ 6 ] had 70% with nine pollen types with the BOW technique, and [ 8 ] obtained 90% in classification of 17 pollen types with neural network.…”

Section: Discussionmentioning

confidence: 99%

“…The system that they developed obtained an accuracy of 89% in the classification of three pollen types, which is a markedly higher accuracy than achieved by a palynologist in routine analysis. In a research study described previously [ 4 ], the researchers reached an overall accuracy of 94% using a system that was developed based on color features to recognize the four most frequent pollen types from Spanish plants with the aim of authenticating the origin of pollen to prevent honey labeling fraud.…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Feature Extraction and Machine Learning for the Classification of Brazilian Savannah Pollen Grains

et al. 2016

View full text Add to dashboard Cite

The classification of pollen species and types is an important task in many areas like forensic palynology, archaeological palynology and melissopalynology. This paper presents the first annotated image dataset for the Brazilian Savannah pollen types that can be used to train and test computer vision based automatic pollen classifiers. A first baseline human and computer performance for this dataset has been established using 805 pollen images of 23 pollen types. In order to access the computer performance, a combination of three feature extractors and four machine learning techniques has been implemented, fine tuned and tested. The results of these tests are also presented in this paper.

show abstract

“…The pollen images of C . pachystachya had 31 hits (85.57%) with this technique, a percentage that is very close (89%) to that obtained by [ 4 ] in a study on Urticaceae pollen grains.…”

Section: Resultssupporting

confidence: 85%

Section: Discussionmentioning

confidence: 71%

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Feature Extraction and Machine Learning for the Classification of Brazilian Savannah Pollen Grains

et al. 2016

View full text Add to dashboard Cite

show abstract

“…Distance‐based one‐class classifiers are based on the assumption that normal data instances occur in dense neighborhoods, while anomalies occur far from their closest neighbors. kNN, originally provided by Dasarathy (1991), is the best‐known distance classifier. The basics of the algorithm for one‐class classification is that the anomaly score of a data instance is defined as the distance with its k th nearest neighbor in a given dataset.…”

Section: Methodsmentioning

confidence: 99%

Silica biotemplating by self‐assembling peptides via serine residues activated by the peptide amino terminal group

Kasotakis

Mitraki

2012

Biopolymers

View full text Add to dashboard Cite

Self-assembling biological materials increasingly serve as templates for the binding of inorganic materials and fabrication of composite nanowires, tubes, etc. with important applications in nanobiotechnology. We have previously reported the use of a self-assembling octapeptide building block as scaffold for the systematic introduction of metal-binding residues, namely cysteines, at the first two amino acids within the sequence (Kasotakis et al., Biopolymers 2009, 92, 164-172). We have also reported unexpected behavior of serine within the octapeptide NH₂ − NSGAITIG − CONH₂( (Asparagine-Serine-Glycine-Alanine-Isoleucine-Threonine-Isoleucine-Glycine) in nucleating gold and platinum nanoparticles. Herein, we report that this serine residue is instrumental in nucleating silica nanoparticles on the surface of the self-assembled fibrils from TEOS (tetraethyl orthosilicate) precursors. We carried out a systematic investigation of the adjacent functionalities and we propose that this serine residue is rendered abnormally nucleophilic through proton abstraction by the N-terminal amino group of the peptide. Peptides with a threonine or a cysteine residue at position 2 are also able to nucleate silica nanoparticles. We propose that rationally designed self-assembling peptides bearing hydroxyl groups adjacent to free amine functionalities could be used for targeted templating of biogenic and even nonbiogenic oxides.

show abstract

Authentication of bee pollen grains in bright‐field microscopy by combining one‐class classification techniques and image processing

Chica

2012

Microscopy Res & Technique

View full text Add to dashboard Cite

A novel method for authenticating pollen grains in bright-field microscopic images is presented in this work. The usage of this new method is clear in many application fields such as bee-keeping sector, where laboratory experts need to identify fraudulent bee pollen samples against local known pollen types. Our system is based on image processing and one-class classification to reject unknown pollen grain objects. The latter classification technique allows us to tackle the major difficulty of the problem, the existence of many possible fraudulent pollen types, and the impossibility of modeling all of them. Different one-class classification paradigms are compared to study the most suitable technique for solving the problem. In addition, feature selection algorithms are applied to reduce the complexity and increase the accuracy of the models. For each local pollen type, a one-class classifier is trained and aggregated into a multiclassifier model. This multiclassification scheme combines the output of all the one-class classifiers in a unique final response. The proposed method is validated by authenticating pollen grains belonging to different Spanish bee pollen types. The overall accuracy of the system on classifying fraudulent microscopic pollen grain objects is 92.3%. The system is able to rapidly reject pollen grains, which belong to nonlocal pollen types, reducing the laboratory work and effort. The number of possible applications of this authentication method in the microscopy research field is unlimited.

show abstract

Discernment of bee pollen loads using computer vision and one-class classification techniques

Cited by 19 publications

References 35 publications

Feature Extraction and Machine Learning for the Classification of Brazilian Savannah Pollen Grains

Feature Extraction and Machine Learning for the Classification of Brazilian Savannah Pollen Grains

Silica biotemplating by self‐assembling peptides via serine residues activated by the peptide amino terminal group

Authentication of bee pollen grains in bright‐field microscopy by combining one‐class classification techniques and image processing

Contact Info

Product

Resources

About