1
|
Fu L, Li S, Kong S, Ni R, Pang H, Sun Y, Hu T, Mu Y, Guo Y, Gong H. Lightweight individual cow identification based on Ghost combined with attention mechanism. PLoS One 2022; 17:e0275435. [PMID: 36201486 PMCID: PMC9536640 DOI: 10.1371/journal.pone.0275435] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/31/2022] [Accepted: 09/15/2022] [Indexed: 11/18/2022] Open
Abstract
Individual cow identification is a prerequisite for intelligent dairy farming management, and is important for achieving accurate and informative dairy farming. Computer vision-based approaches are widely considered because of their non-contact and practical advantages. In this study, a method based on the combination of Ghost and attention mechanism is proposed to improve ReNet50 to achieve non-contact individual recognition of cows. In the model, coarse-grained features of cows are extracted using a large sensory field of cavity convolution, while reducing the number of model parameters to some extent. ResNet50 consists of two Bottlenecks with different structures, and a plug-and-play Ghost module is inserted between the two Bottlenecks to reduce the number of parameters and computation of the model using common linear operations without reducing the feature map. In addition, the convolutional block attention module (CBAM) is introduced after each stage of the model to help the model to give different weights to each part of the input and extract the more critical and important information. In our experiments, a total of 13 cows' side view images were collected to train the model, and the final recognition accuracy of the model was 98.58%, which was 4.8 percentage points better than the recognition accuracy of the original ResNet50, the number of model parameters was reduced by 24.85 times, and the model size was only 3.61 MB. In addition, to verify the validity of the model, it is compared with other networks and the results show that our model has good robustness. This research overcomes the shortcomings of traditional recognition methods that require human extraction of features, and provides theoretical references for further animal recognition.
Collapse
Affiliation(s)
- Lili Fu
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Shijun Li
- College of Electronic and Information Engineering, Wuzhou University, Wuzhou, China
| | - Shuolin Kong
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Ruiwen Ni
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Haohong Pang
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Yu Sun
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Tianli Hu
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Ye Mu
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Ying Guo
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - He Gong
- College of Information Technology, Jilin Agricultural University, Changchun, China
- * E-mail:
| |
Collapse
|
2
|
Li S, Fu L, Sun Y, Mu Y, Chen L, Li J, Gong H. Individual dairy cow identification based on lightweight convolutional neural network. PLoS One 2021; 16:e0260510. [PMID: 34843562 PMCID: PMC8629223 DOI: 10.1371/journal.pone.0260510] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/30/2021] [Accepted: 11/09/2021] [Indexed: 11/18/2022] Open
Abstract
In actual farms, individual livestock identification technology relies on large models with slow recognition speeds, which seriously restricts its practical application. In this study, we use deep learning to recognize the features of individual cows. Alexnet is used as a skeleton network for a lightweight convolutional neural network that can recognise individual cows in images with complex backgrounds. The model is improved for multiple multiscale convolutions of Alexnet using the short-circuit connected BasicBlock to fit the desired values and avoid gradient disappearance or explosion. An improved inception module and attention mechanism are added to extract features at multiple scales to enhance the detection of feature points. In experiments, side-view images of 13 cows were collected. The proposed method achieved 97.95% accuracy in cow identification with a single training time of only 6 s, which is one-sixth that of the original Alexnet. To verify the validity of the model, the dataset and experimental parameters were kept constant and compared with the results of Vgg16, Resnet50, Mobilnet V2 and GoogLenet. The proposed model ensured high accuracy while having the smallest parameter size of 6.51 MB, which is 1.3 times less than that of the Mobilnet V2 network, which is famous for its light weight. This method overcomes the defects of traditional methods, which require artificial extraction of features, are often not robust enough, have slow recognition speeds, and require large numbers of parameters in the recognition model. The proposed method works with images with complex backgrounds, making it suitable for actual farming environments. It also provides a reference for the identification of individual cows in images with complex backgrounds.
Collapse
Affiliation(s)
- Shijun Li
- College of Electronic and Information Engineering, Wuzhou University, Wuzhou, China
| | - Lili Fu
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Yu Sun
- College of Information Technology, Jilin Agricultural University, Changchun, China
- Jilin Province Agricultural Internet of Things Technology Collaborative Innovation Center, Changchun, China
- Jilin Province Intelligent Environmental Engineering Research Center, Changchun, China
- Jilin Province Information Technology and Intelligent Agricultural Engineering Research Center, Changchun, China
| | - Ye Mu
- College of Information Technology, Jilin Agricultural University, Changchun, China
- Jilin Province Agricultural Internet of Things Technology Collaborative Innovation Center, Changchun, China
- Jilin Province Intelligent Environmental Engineering Research Center, Changchun, China
- Jilin Province Information Technology and Intelligent Agricultural Engineering Research Center, Changchun, China
| | - Lin Chen
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - Ji Li
- College of Information Technology, Jilin Agricultural University, Changchun, China
| | - He Gong
- College of Information Technology, Jilin Agricultural University, Changchun, China
- Jilin Province Agricultural Internet of Things Technology Collaborative Innovation Center, Changchun, China
- Jilin Province Intelligent Environmental Engineering Research Center, Changchun, China
- Jilin Province Information Technology and Intelligent Agricultural Engineering Research Center, Changchun, China
| |
Collapse
|
3
|
Caja G, Díaz-Medina E, Salama AAK, Salama OAE, El-Shafie MH, El-Metwaly HA, Ayadi M, Aljumaah RS, Alshaikh MA, Yahyahoui MH, Seddik MM, Hammadi M, Khorchani T, Amann O, Cabrera S. Comparison of visual and electronic devices for individual identification of dromedary camels under different farming conditions. J Anim Sci 2016; 94:3561-3571. [PMID: 27695805 PMCID: PMC7199663 DOI: 10.2527/jas.2016-0472] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/13/2022] Open
Abstract
The camel industry uses traditional (i.e., iron brands and ear tags) and modern (i.e., microchips) identification (ID) systems without having performance results of reference. Previously iron-branded ( = 45; 1 yr) and microchipped ( = 59; 7 yr) camels showed problems of healing (8.6% of brands) and reading (only 42.9% of brands and 69.5% of microchips were readable), which made their use inadvisable. With the aim of proposing suitable ID systems for different farming conditions, an on-field study was performed using a total of 528 dromedaries at 4 different locations (Egypt, = 83; Spain, = 304; Saudi Arabia, = 90; and Tunisia, = 51). The ID devices tested were visual (button ear tags, 28.5 mm diameter, = 178; double flag ear tags, 50 by 15 mm, = 83; both made of polyurethane) and electronic (ear tags, = 90, and rumen boluses, = 555). Electronic ear tags were polyurethane-loop type (75 by 9 mm) with a container in which a 22-mm transponder of full-duplex technology was lodged. Electronic boluses of 7 types, varying in dimensions (50 to 76 mm length, 11 to 21 mm width, and 12.7 to 82.1 g weight) and specific gravity (SG; 1.49 to 3.86) and each of them containing a 31-mm transponder of half-duplex technology, were all administered to the dromedaries at the beginning of the study. When a low-SG bolus was lost, a high-SG bolus was readministered. Readability rates of each ID system were evaluated during 1 to 3 yr, according to device and location, and yearly values were estimated for comparison. On a yearly basis, visual ear tag readability was not fully satisfactory; it was lower for rectangular ear tags (66.3%) than for button ear tags (80.9%). Yearly readability of electronic ear tags was 93.7%. Bolus readability dramatically varied according to their SG; the SG < 2.0 boluses were fully lost after 8 mo. In contrast, the SG > 3.0 boluses were efficiently retained (99.6 to 100%) at all locations. In conclusion, according to the expected long lifespan of camels, low ID performances were observed for iron brands, injectable microchips, and ear tags (visual and electronic), making their use inadvisable as unique ID systems in camels. The high readability of dense electronic boluses recommended their use as a permanent ID device of reference in camels.
Collapse
Affiliation(s)
- G. Caja
- Group of Ruminant Research (G2R), Department of Animal and Food Sciences, Universitat Autònoma de Barcelona, Bellaterra, Spain
- Corresponding author:
| | - E. Díaz-Medina
- Group of Ruminant Research (G2R), Department of Animal and Food Sciences, Universitat Autònoma de Barcelona, Bellaterra, Spain
- Oasis Park-Museo del Campo Majorero, La Lajita, Fuerteventura, Las Palmas de Gran Canaria, Spain
| | - A. A. K. Salama
- Group of Ruminant Research (G2R), Department of Animal and Food Sciences, Universitat Autònoma de Barcelona, Bellaterra, Spain
- Animal Production Research Institute (APRI), Agricultural Research Center, Dokki, Giza, Egypt
| | - O. A. E. Salama
- Animal Production Research Institute (APRI), Agricultural Research Center, Dokki, Giza, Egypt
| | - M. H. El-Shafie
- Animal Production Research Institute (APRI), Agricultural Research Center, Dokki, Giza, Egypt
| | - H. A. El-Metwaly
- Animal Production Research Institute (APRI), Agricultural Research Center, Dokki, Giza, Egypt
| | - M. Ayadi
- Department of Animal Production, College of Food and Agriculture Sciences, King Saud University (KSU), Riyadh, Saudi Arabia
- Département de Biotechnologie Animale, Institut Supérieur de Biotechnologie de Beja, Université de Jendouba, Tunisia
| | - R. S. Aljumaah
- Department of Animal Production, College of Food and Agriculture Sciences, King Saud University (KSU), Riyadh, Saudi Arabia
| | - M. A. Alshaikh
- Department of Animal Production, College of Food and Agriculture Sciences, King Saud University (KSU), Riyadh, Saudi Arabia
| | - M. H. Yahyahoui
- Département de Biotechnologie Animale, Institut Supérieur de Biotechnologie de Beja, Université de Jendouba, Tunisia
| | - M. M. Seddik
- Département de Biotechnologie Animale, Institut Supérieur de Biotechnologie de Beja, Université de Jendouba, Tunisia
| | - M. Hammadi
- Département de Biotechnologie Animale, Institut Supérieur de Biotechnologie de Beja, Université de Jendouba, Tunisia
| | - T. Khorchani
- Département de Biotechnologie Animale, Institut Supérieur de Biotechnologie de Beja, Université de Jendouba, Tunisia
| | - O. Amann
- Oasis Park-Museo del Campo Majorero, La Lajita, Fuerteventura, Las Palmas de Gran Canaria, Spain
| | - S. Cabrera
- Oasis Park-Museo del Campo Majorero, La Lajita, Fuerteventura, Las Palmas de Gran Canaria, Spain
| |
Collapse
|
4
|
Karakus F, Demir AÖ, Akkol S, Düzgün A, Karakus M. Performance of electronic and visual ear tags in lambs under extensive conditions in Turkey. Arch Anim Breed 2015. [DOI: 10.5194/aab-58-287-2015] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/11/2022] Open
Abstract
Abstract. The objective of this study was to evaluate the effectiveness of electronic and visual ear tags in animal traceability, and to investigate the effect of placement site on ear-tag retention in Akkaraman lambs under rural conditions. A total of 380 lambs were identified with electronic and visual ear tags. Electronic and visual ear tags displayed 98.9 and 98.7 % readability at the end of 7 months, and 98.0 and 98.0 % readability at the end of the first year after tagging, respectively. Regarding the placement site, it was observed that there was more loss in ear tags placed on the mid-point part of the ear than the first-quarter part from the head side, but the difference was not statistically significant (P > 0.05). Breakages and electronic failures were not recorded during this study. In conclusion, electronic and visual ear tags demonstrated similar on-farm efficiency for the identification of Akkaraman lambs and fulfilled the minimum efficiency of 98 % required by the International Committee for Animal Recording (ICAR) for an official animal identification device at the end of the first year after tagging. Based on the findings of the study, placement of the ear tag in a cranial position and near the base of the ear would be advised.
Collapse
|
5
|
Caja G, Carné S, Salama A, Ait-Saidi A, Rojas-Olivares M, Rovai M, Capote J, Castro N, Argüello A, Ayadi M, Aljumaah R, Alshaikh M. State-of-the-art of electronic identification techniques and applications in goats. Small Rumin Res 2014. [DOI: 10.1016/j.smallrumres.2014.05.012] [Citation(s) in RCA: 11] [Impact Index Per Article: 1.1] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
|
6
|
Affiliation(s)
- A. Argüello
- a Department of Animal Science , Universidad de Las Palmas de Gran Canaria , Arucas, Las Palmas, Spain
| |
Collapse
|
7
|
Carné S, Caja G, Ghirardi J, Salama A. Modeling the retention of rumen boluses for the electronic identification of goats. J Dairy Sci 2011; 94:716-26. [DOI: 10.3168/jds.2010-3210] [Citation(s) in RCA: 8] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/26/2010] [Accepted: 08/30/2010] [Indexed: 11/19/2022]
|