Limits...
Species Identification of Food Contaminating Beetles by Recognizing Patterns in Microscopic Images of Elytra Fragments.

Park SI, Bisgin H, Ding H, Semey HG, Langley DA, Tong W, Xu J - PLoS ONE (2016)

Bottom Line: A crucial step of food contamination inspection is identifying the species of beetle fragments found in the sample, since the presence of some storage beetles is a good indicator of insanitation or potential food safety hazards.Both global and local characteristics were quantified and used as feature inputs to artificial neural networks for species classification.Through examining the overall and per species accuracies, we further demonstrated that the local features are better suited than the global features for species identification.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science, Texas A&M University, College Station, Texas, United States of America.

ABSTRACT
A crucial step of food contamination inspection is identifying the species of beetle fragments found in the sample, since the presence of some storage beetles is a good indicator of insanitation or potential food safety hazards. The current pratice, visual examination by human analysts, is time consuming and requires several years of experience. Here we developed a species identification algorithm which utilizes images of microscopic elytra fragments. The elytra, or hardened forewings, occupy a large portion of the body, and contain distinctive patterns. In addition, elytra fragments are more commonly recovered from processed food products than other body parts due to their hardness. As a preliminary effort, we chose 15 storage product beetle species frequently detected in food inspection. The elytra were then separated from the specimens and imaged under a microscope. Both global and local characteristics were quantified and used as feature inputs to artificial neural networks for species classification. With leave-one-out cross validation, we achieved overall accuracy of 80% through the proposed global and local features, which indicates that our proposed features could differentiate these species. Through examining the overall and per species accuracies, we further demonstrated that the local features are better suited than the global features for species identification. Future work will include robust testing with more beetle species and algorithm refinement for a higher accuracy.

No MeSH data available.


Related in: MedlinePlus

Identification accuracy per species.One panel plots the results for each feature set: (A) a subset of global features (i.e., “global features 1”), (B) the second subset of global features with hair/hole/line features (i.e., “global features 2”), (C) the set of local features (i.e., “local features”), (D) all features (i.e., “global and local features”). Each bar plots the mean and standard deviation of identification accuracy achieved during 100 rounds of cross validation.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4920424&req=5

pone.0157940.g012: Identification accuracy per species.One panel plots the results for each feature set: (A) a subset of global features (i.e., “global features 1”), (B) the second subset of global features with hair/hole/line features (i.e., “global features 2”), (C) the set of local features (i.e., “local features”), (D) all features (i.e., “global and local features”). Each bar plots the mean and standard deviation of identification accuracy achieved during 100 rounds of cross validation.

Mentions: Along with the overall performance, we examined identification accuracies for each species. Each subpanel of Fig 12 shows per species accuracies obtained by the global features 1 or 2, local features, or all features. The statistical, spectral, color, and angular edge response features in the “global features 1” were capable of identifying many species, but they were not sufficient to distinguish species with a high resemblance. Even with hair/hole/line quantificaiton, the “global featuers 2” rendered a meager performance improvement. The better performance of the local features demonstrated that the segmented local attributes represented in the local features were most effective for characterizing individual species that look alike. Besides a small improvement in overall performance, the inclusion of the global features led to more balanced per species performance than the local features alone. The lowest per species mean accuracy was 66% (for Species 12) by all features but only 52% (for Species 8) by the local features. With all features, 5 species (Species 1, 3, 9, 11, and 15) were identified with high accuracy (≥ 85%) and 5 species (Species 2, 4, 5, 6, and 7) were well-classified (75%–85%). Species 8, 10, 12, 13, and 14 were recognized with decent accuracy (65%–75%). Importantly, Species 12, 13, and 14 were classified correctly at genus level due to their very similar characteristics as mentioned earlier. Thus, we confirmed that some objects of interest such as hairs, holes and lines were crucial features for species identification.


Species Identification of Food Contaminating Beetles by Recognizing Patterns in Microscopic Images of Elytra Fragments.

Park SI, Bisgin H, Ding H, Semey HG, Langley DA, Tong W, Xu J - PLoS ONE (2016)

Identification accuracy per species.One panel plots the results for each feature set: (A) a subset of global features (i.e., “global features 1”), (B) the second subset of global features with hair/hole/line features (i.e., “global features 2”), (C) the set of local features (i.e., “local features”), (D) all features (i.e., “global and local features”). Each bar plots the mean and standard deviation of identification accuracy achieved during 100 rounds of cross validation.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4920424&req=5

pone.0157940.g012: Identification accuracy per species.One panel plots the results for each feature set: (A) a subset of global features (i.e., “global features 1”), (B) the second subset of global features with hair/hole/line features (i.e., “global features 2”), (C) the set of local features (i.e., “local features”), (D) all features (i.e., “global and local features”). Each bar plots the mean and standard deviation of identification accuracy achieved during 100 rounds of cross validation.
Mentions: Along with the overall performance, we examined identification accuracies for each species. Each subpanel of Fig 12 shows per species accuracies obtained by the global features 1 or 2, local features, or all features. The statistical, spectral, color, and angular edge response features in the “global features 1” were capable of identifying many species, but they were not sufficient to distinguish species with a high resemblance. Even with hair/hole/line quantificaiton, the “global featuers 2” rendered a meager performance improvement. The better performance of the local features demonstrated that the segmented local attributes represented in the local features were most effective for characterizing individual species that look alike. Besides a small improvement in overall performance, the inclusion of the global features led to more balanced per species performance than the local features alone. The lowest per species mean accuracy was 66% (for Species 12) by all features but only 52% (for Species 8) by the local features. With all features, 5 species (Species 1, 3, 9, 11, and 15) were identified with high accuracy (≥ 85%) and 5 species (Species 2, 4, 5, 6, and 7) were well-classified (75%–85%). Species 8, 10, 12, 13, and 14 were recognized with decent accuracy (65%–75%). Importantly, Species 12, 13, and 14 were classified correctly at genus level due to their very similar characteristics as mentioned earlier. Thus, we confirmed that some objects of interest such as hairs, holes and lines were crucial features for species identification.

Bottom Line: A crucial step of food contamination inspection is identifying the species of beetle fragments found in the sample, since the presence of some storage beetles is a good indicator of insanitation or potential food safety hazards.Both global and local characteristics were quantified and used as feature inputs to artificial neural networks for species classification.Through examining the overall and per species accuracies, we further demonstrated that the local features are better suited than the global features for species identification.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science, Texas A&M University, College Station, Texas, United States of America.

ABSTRACT
A crucial step of food contamination inspection is identifying the species of beetle fragments found in the sample, since the presence of some storage beetles is a good indicator of insanitation or potential food safety hazards. The current pratice, visual examination by human analysts, is time consuming and requires several years of experience. Here we developed a species identification algorithm which utilizes images of microscopic elytra fragments. The elytra, or hardened forewings, occupy a large portion of the body, and contain distinctive patterns. In addition, elytra fragments are more commonly recovered from processed food products than other body parts due to their hardness. As a preliminary effort, we chose 15 storage product beetle species frequently detected in food inspection. The elytra were then separated from the specimens and imaged under a microscope. Both global and local characteristics were quantified and used as feature inputs to artificial neural networks for species classification. With leave-one-out cross validation, we achieved overall accuracy of 80% through the proposed global and local features, which indicates that our proposed features could differentiate these species. Through examining the overall and per species accuracies, we further demonstrated that the local features are better suited than the global features for species identification. Future work will include robust testing with more beetle species and algorithm refinement for a higher accuracy.

No MeSH data available.


Related in: MedlinePlus