Automated pollen identification system for forensic geo-historical location applications

Hwang, Grace M.; Riley, Kim C.; Christou, C.T.; Jacyna, Garry M.; Woodard, Jeffrey P.; Ryan, Regina M.; Punyasena, Surangi W.; Bush, Mark B.; Valencia, Bryan G.; McMichael, Crystal N.H.; Masters, David

doi:10.1109/ths.2013.6699018

Cited by 6 publications

(4 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One can use this trace evidence to attribute the possible location history of the objects recovered at a scene or from the suspects of interest related to the scene (Bryant & Jones, 2006; Mildenhall et al, 2006). Pollen grains and fungal spores are suitable for trace evidence because of their environmental ubiquity and durability in most terrestrial environments around the globe (Grantham et al, 2015; Helderop et al, 2021; Hwang et al, 2013; Tong et al, 2021). One can recover these particulate materials from external or internal human bodies or items contacted by humans or animals, including computers left at crime scenes and digging tools at burial sites.…”

Section: Introductionmentioning

confidence: 99%

Using habitat suitability models for multiscale forensic geolocation analysis

Wang

Miller

Grubesic

et al. 2023

Transactions in GIS

View full text Add to dashboard Cite

Pollen is one of the most durable environmental materials that law enforcement agencies recover as trace evidence from people and objects. Although links between objects and geographic locations are essential during legal investigations, the approach of using pollen and other microbial fingerprints to build these links in an analytical framework is still underutilized. This study uses bees as objects that are mobile and collects environmental traces as a test case to determine the efficacy of predictive geolocation efforts with recovered pollen and species distribution models at both subcontinental and global scales. Results demonstrate promising performance in both the predictive capability of species distribution models and identification of possible location history of bees at both study extents. When coupling pollen with other categories of evidentiary items, this geographic attribution framework can aid law enforcement personnel in refining investigation priorities and optimizing search strategies.

show abstract

Section: Introductionmentioning

confidence: 99%

Using habitat suitability models for multiscale forensic geolocation analysis

Wang

Miller

Grubesic

et al. 2023

Transactions in GIS

View full text Add to dashboard Cite

show abstract

“…A palinologiaé a ciência que estuda os grãos de pólen e esporos, estes são objetos microscópicos produzidos pelas plantas classificadas como: briófitas, pteridófitas, gimnospermas e angiospermas, ou seja, são células reprodutivas provenientes das flores possuindo um tamanho médio de 10 a 50 µmetros [Joosten and de Klerk, 2002]. Ademais, outraś areas de conhecimento ramificaram-se da palinologia, como o estudo de polens de mel através da melissopalinologia [Von Der Ohe et al, 2004], o estudo de polens causadores de alergia na imunologia [Delves et al, 2017] e o estudo de ecossistemas do passado através da paleoecologia [Jackson and Williams, 2004].…”

Section: Introductionunclassified

“…A atividade de classificar e contar grãos de pólen pode ser automatizada com o uso da visão computacional e da aprendizagem de máquina [Hwang et al, 2013]. De acordo com [Prince, 2012], o objetivo da visão computacionalé desenvolver computacionalmente a habilidade visual que os seres humanos já possuem e tornar a tarefa mais rápida e eficiente, utilizando-se de algoritmos e técnicas computacionais desenvolvidas para determinado objetivo.…”

Section: Introductionunclassified

A Comparison Study of Deep Convolutional Neural Networks for the Classification of Brazilian Savannah Pollen Grains: Preliminary Results

Aristimunha

Borges

Gonçalves

et al. 2019

Anais Do XV Workshop De Visão Computacional (WVC 2019)

View full text Add to dashboard Cite

The classiﬁcation of pollen grains images are currently done manually and visually, being a weariful task and predisposed to mistakes due to human exhaustion. In this paper, the authors introduce an automatic classiﬁcation of 55 different pollen grain classes, using convolutional neural networks. Different architectures and hyperparameters were used to improve the classiﬁcation result. Using the networks VGG16, VGG19, and InceptionV3, were obtained accuracy rates over 93.58%.

show abstract

“…In the work [2] there is described a semiautomatic system for pollen grains analysis with the usage of computer vision for increasing the number of processed data. Computer vision algorithms was used for automation.…”

Section: Introductionmentioning

confidence: 99%

The use of convolutional neural network LeNet for pollen grains classification

Korobeynikov¹,

Kamalova²,

Palabugin³

et al. 2018

Instrumentation Engineering, Electronics and Telecommunications

View full text Add to dashboard Cite

The convolutional neural network from the LenetMnistExample of the DeepLearning4j framework is described and applied for pollen grains classification.The selected basic topology of the LeNet neural network was not changed; the loading of images was modified, and the number of classes of the classification task (outputNum) as well as the subsample size of examples (batchSize) were changed. Training (1520 photos) and test (380 photos) samples of four classes of pollen grains were formed. The quality metrics values calculated according to the results of the test sample classification are: 1) Accuracy = 0.9289; 2) Precision = 0.9306; 3) Recall = 0.9266; 4) F1 Score = 0.9282.

show abstract

Automated pollen identification system for forensic geo-historical location applications

Cited by 6 publications

References 11 publications

Using habitat suitability models for multiscale forensic geolocation analysis

Using habitat suitability models for multiscale forensic geolocation analysis

A Comparison Study of Deep Convolutional Neural Networks for the Classification of Brazilian Savannah Pollen Grains: Preliminary Results

The use of convolutional neural network LeNet for pollen grains classification

Contact Info

Product

Resources

About