Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification

Abstract Background The correct identification of pills is very important to ensure the safe administration of drugs to patients. Here, we use three current mainstream object detection models, namely RetinaNet, Single Shot Multi-Box Detector (SSD), and You Only Look Once v3(YOLO v3), to identify pil...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Lu Tan, Tianran Huangfu, Liyao Wu, Wenying Chen
Formato:	article
Lenguaje:	EN
Publicado:	BMC 2021
Materias:	Convolutional neural network RetinaNet SSD YOLO v3 Pill identification Computer applications to medicine. Medical informatics R858-859.7
Acceso en línea:	https://doaj.org/article/b8730b5e504e4900ae76aa43042b7a96
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:b8730b5e504e4900ae76aa43042b7a96
record_format	dspace
spelling	oai:doaj.org-article:b8730b5e504e4900ae76aa43042b7a962021-11-28T12:26:15ZComparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification10.1186/s12911-021-01691-81472-6947https://doaj.org/article/b8730b5e504e4900ae76aa43042b7a962021-11-01T00:00:00Zhttps://doi.org/10.1186/s12911-021-01691-8https://doaj.org/toc/1472-6947Abstract Background The correct identification of pills is very important to ensure the safe administration of drugs to patients. Here, we use three current mainstream object detection models, namely RetinaNet, Single Shot Multi-Box Detector (SSD), and You Only Look Once v3(YOLO v3), to identify pills and compare the associated performance. Methods In this paper, we introduce the basic principles of three object detection models. We trained each algorithm on a pill image dataset and analyzed the performance of the three models to determine the best pill recognition model. The models were then used to detect difficult samples and we compared the results. Results The mean average precision (MAP) of RetinaNet reached 82.89%, but the frames per second (FPS) is only one third of YOLO v3, which makes it difficult to achieve real-time performance. SSD does not perform as well on the indicators of MAP and FPS. Although the MAP of YOLO v3 is slightly lower than the others (80.69%), it has a significant advantage in terms of detection speed. YOLO v3 also performed better when tasked with hard sample detection, and therefore the model is more suitable for deployment in hospital equipment. Conclusion Our study reveals that object detection can be applied for real-time pill identification in a hospital pharmacy, and YOLO v3 exhibits an advantage in detection speed while maintaining a satisfactory MAP.Lu TanTianran HuangfuLiyao WuWenying ChenBMCarticleConvolutional neural networkRetinaNetSSDYOLO v3Pill identificationComputer applications to medicine. Medical informaticsR858-859.7ENBMC Medical Informatics and Decision Making, Vol 21, Iss 1, Pp 1-11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Convolutional neural network RetinaNet SSD YOLO v3 Pill identification Computer applications to medicine. Medical informatics R858-859.7
spellingShingle	Convolutional neural network RetinaNet SSD YOLO v3 Pill identification Computer applications to medicine. Medical informatics R858-859.7 Lu Tan Tianran Huangfu Liyao Wu Wenying Chen Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification
description	Abstract Background The correct identification of pills is very important to ensure the safe administration of drugs to patients. Here, we use three current mainstream object detection models, namely RetinaNet, Single Shot Multi-Box Detector (SSD), and You Only Look Once v3(YOLO v3), to identify pills and compare the associated performance. Methods In this paper, we introduce the basic principles of three object detection models. We trained each algorithm on a pill image dataset and analyzed the performance of the three models to determine the best pill recognition model. The models were then used to detect difficult samples and we compared the results. Results The mean average precision (MAP) of RetinaNet reached 82.89%, but the frames per second (FPS) is only one third of YOLO v3, which makes it difficult to achieve real-time performance. SSD does not perform as well on the indicators of MAP and FPS. Although the MAP of YOLO v3 is slightly lower than the others (80.69%), it has a significant advantage in terms of detection speed. YOLO v3 also performed better when tasked with hard sample detection, and therefore the model is more suitable for deployment in hospital equipment. Conclusion Our study reveals that object detection can be applied for real-time pill identification in a hospital pharmacy, and YOLO v3 exhibits an advantage in detection speed while maintaining a satisfactory MAP.
format	article
author	Lu Tan Tianran Huangfu Liyao Wu Wenying Chen
author_facet	Lu Tan Tianran Huangfu Liyao Wu Wenying Chen
author_sort	Lu Tan
title	Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification
title_short	Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification
title_full	Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification
title_fullStr	Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification
title_full_unstemmed	Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification
title_sort	comparison of retinanet, ssd, and yolo v3 for real-time pill identification
publisher	BMC
publishDate	2021
url	https://doaj.org/article/b8730b5e504e4900ae76aa43042b7a96
work_keys_str_mv	AT lutan comparisonofretinanetssdandyolov3forrealtimepillidentification AT tianranhuangfu comparisonofretinanetssdandyolov3forrealtimepillidentification AT liyaowu comparisonofretinanetssdandyolov3forrealtimepillidentification AT wenyingchen comparisonofretinanetssdandyolov3forrealtimepillidentification
_version_	1718407977970434048

Comparison of RetinaNet, SSD, and YOLO v3 for real-time pill identification

Ejemplares similares