Accesso libero

Detection of pig based on improved RESNET model in natural scene

Weixian Song

Weixian Song

College of Electrical and Information, Northeast Agricultural UniversityHarbin, China
Cerca questo autore su
Sciendo|Google Scholar
Song, Weixian
, 
Junlong Fang

Junlong Fang

College of Electrical and Information, Northeast Agricultural UniversityHarbin, China
Cerca questo autore su
Sciendo|Google Scholar
Fang, Junlong
, 
Runtao Wang

Runtao Wang

College of Electrical and Information, Northeast Agricultural UniversityHarbin, China
Cerca questo autore su
Sciendo|Google Scholar
Wang, Runtao
 e 
Kezhu Tan

Kezhu Tan

College of Electrical and Information, Northeast Agricultural UniversityHarbin, China
Cerca questo autore su
Sciendo|Google Scholar
Tan, Kezhu
  
30 ott 2021
Detection of pig based on improved RESNET model in natural scene's Cover Image
INFORMAZIONI SU QUESTO ARTICOLO
Articolo precedente
Articolo Successivo
Cita
Scarica la copertina
Pubblicato online: 30 ott 2021
Pagine: 215 - 226
Ricevuto: 26 mar 2021
Accettato: 29 giu 2021
DOI: https://doi.org/10.2478/amns.2021.2.00040
Parole chiave
© 2021 Weixian Song et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.

Fig. 1

Part dataset of this work.
Part dataset of this work.

Fig. 2

Residual module.
Residual module.

Fig. 3

Residual block structure A.
Residual block structure A.

Fig. 4

Residual block structure B.
Residual block structure B.

Fig. 5

The overall structure of the improved model.
The overall structure of the improved model.

Fig. 6

The results of multiple pig detection.
The results of multiple pig detection.

Fig. 7

The results of individual pig detection.
The results of individual pig detection.

Fig. 8

Rotation of driving wheels around the pipe axis.
Rotation of driving wheels around the pipe axis.

Fig. 9

The training and testing results of the ResNet model. (a) Dataset 1 and (b) Dataset 2.
The training and testing results of the ResNet model. (a) Dataset 1 and (b) Dataset 2.

Fig. 10

Experimental results of the original AlexNet model. (a) The relationship between test accuracy and the number of iterations and (b) The relationship between model loss and iterations..
Experimental results of the original AlexNet model. (a) The relationship between test accuracy and the number of iterations and (b) The relationship between model loss and iterations..

Fig. 11

The training and testing results of GoogLeNet model.
The training and testing results of GoogLeNet model.

Fig. 12

The training and testing results of ResNet model.
The training and testing results of ResNet model.

Fig. 13

The training and test results of the improved model.
The training and test results of the improved model.

Training parameter settings of this study_

Parameter name test_iter test_interval base_lr weight_decay gamma max_iter solver_mode
Parameter setting 120 500 0.1 0.0001 0.1 50,000 GPU

The training and testing results of dataset2 of different CNN models_

CNN model Training accuracy (%) Testing accuracy (%) Loss (%)
AlexNet 94.2 90.4 22.1
GoogLeNet 96.7 93.1 16.7
ResNet 97.7 95.3 10.7
Improved ResNet 98.2 96.4 7.7

Distribution of training and testing images for dataset 1_

Pig posture The number of original images The number of images after expanding The images in the training set The images in the testing set
Standing 600 1200 983 217
Standing with lowering the head 200 400 316 84
Raising head 240 480 390 90
Laying down 55 110 88 22

Distribution of training and testing images for dataset 2_

Pig posture The number of original images The number of images after expanding The images in the training set The images in the testing set
Standing 600 4800 3844 956
Standing with lowering the head 200 1600 1264 336
Raising head 240 1920 1560 360
Laying down 55 440 352 88
Lingua:
Inglese
Frequenza di pubblicazione:
1 volte all'anno
Argomenti della rivista:
Scienze biologiche, Scienze della vita, altro, Matematica, Matematica applicata, Matematica generale, Fisica, Fisica, altro
Feed RSS della rivista