Center clustering network improves piglet counting under occlusion

Huang, Endai; Mao, Axiu; Gan, Haiming; Ceballos, Maria Camila; Parsons, Thomas D.; Xue, Yong; Liu, Kai

doi:10.1016/j.compag.2021.106417

Cited by 20 publications

(17 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The six videos, with a frame rate of 7 FPS and a resolution of 1024×768 pixels, were captured by overhead cameras located 2 m above the farrowing crates. Six datasets, containing 4,600 images in total, were extracted from the six videos and labeled for piglet centers, individual piglet masks, and sow masks (Huang et al, 2021b). Datasets 1, 2, and 3 contained data from pens with half-open farrowing crates, and Datasets 4, 5, and 6 contained data from pens with closed crates (Fig.…”

Section: Datamentioning

confidence: 99%

“…Two types of partial occlusion on piglets were defined: body-separated occlusion (BSO) and part-missing occlusion (PMO) (Huang et al, 2021b). In BSO, the occlusions (e.g., occlusions from farrowing crates) separated a target into two or more parts, whereas in PMO, a terminal part of a target (e.g., the head of a piglet) was occluded.…”

Section: Datamentioning

confidence: 99%

“…Such occlusions have non-negligible consequences for piglet due to piglets' small size (e.g., only the head of a piglet may be visible while the other parts of the animal are occluded by farrowing crates). Huang et al (2021b) found that occlusions occurred in 97.7% of the images collected from farrowing pens with farrowing crates. These unexpected occlusions may decrease the accuracy and even cause the failure of deep-learning-based methods (Huang et al, 2021a).…”

Section: Introductionmentioning

confidence: 95%

“…This study aims to develop a novel occlusionresistant instance segmentation method for piglets in farrowing pens with frequently occurring occlusions. We propose CClusnet-Inseg, a Center Clustering Network for instance segmentation, based on a previous model called Center Clustering Network for counting (CClusnet-Count; (Huang et al, 2021b)). We use every visible pixel to predict an object center, and after these centers are clustered into different groups, we trace these clustered centers to form different masks according to their groups.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Occlusion-Resistant Instance Segmentation of Piglets in Farrowing Pens Using Center Clustering Network

Huang¹,

Mao²,

Wu³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Computer vision enables the development of new approaches to monitor the behavior, health, and welfare of animals. Instance segmentation is a high-precision method in computer vision for detecting individual animals of interest. This method can be used for in-depth analysis of animals, such as examining their subtle interactive behaviors, from videos and images. However, existing deep-learning-based instance segmentation methods have been mostly developed based on public datasets, which largely omit heavy occlusion problems; therefore, these methods have limitations in real-world applications involving object occlusions, such as farrowing pen systems used on pig farms in which the farrowing crates often impede the sow and piglets. In this paper, we propose a novel occlusion-resistant Center Clustering Network for instance segmentation, dubbed as CClusnet-Inseg. Specifically, CClusnet-Inseg uses each pixel to predict object centers and trace these centers to form masks based on clustering results, which consists of a network for segmentation and center offset vector map, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, Centers-to-Mask (C2M) and Remain-Centers-to-Mask (RC2M) algorithms, and a pseudo-occlusion generator (POG). In all, 4,600 images were extracted from six videos collected from six farrowing pens to train and validate our method. CClusnet-Inseg achieves a mean average precision (mAP) of 83.6; it outperformed YOLACT++ and Mask R-CNN, which had mAP values of 81.2 and 74.7, respectively. We conduct comprehensive ablation studies to demonstrate the advantages and effectiveness of core modules of our method. In addition, we apply CClusnet-Inseg to multi-object tracking for animal monitoring, and the predicted object center that is a conjunct output could serve as an occlusion-resistant representation of the location of an object.

show abstract

Section: Datamentioning

confidence: 99%

Section: Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 95%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Occlusion-Resistant Instance Segmentation of Piglets in Farrowing Pens Using Center Clustering Network

Huang¹,

Mao²,

Wu³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…In 2020, Xu et al developed an auto sheep counting system based on multi-object detection, tracking, and extrapolation techniques [ 3 ]. In 2021, to address problems caused by the partial occlusion, a two-stage center clustering network (CClusnet) was developed to improve automated piglet counting performance [ 4 ]. Jensen et al made use of convolutional neural networks with a linear regression output to realize automatic counting and positioning of slaughter pigs within a pen [ 5 ].…”

Section: Introductionmentioning

confidence: 99%

An Integrated Goat Head Detection and Automatic Counting Method Based on Deep Learning

Liu

Chen

et al. 2022

Animals

View full text Add to dashboard Cite

Goat farming is one of the pillar industries for sustainable development of national economies in some countries and plays an active role in social and economic development. In order to realize the precision and intelligence of goat breeding, this paper describes an integrated goat detection and counting method based on deep learning. First, we constructed a new dataset of video images of goats for the object tracking task. Then, we took YOLOv5 as the baseline of the object detector and improved it using a series of advanced methods, including: using RandAugment to explore suitable data augmentation strategies in a real goat barn environment, using AF-FPN to improve the network’s ability to represent multi-scale objects, and using the Dynamic Head framework to unify the attention mechanism with the detector’s heads to improve its performance. The improved detector achieved 92.19% mAP, a significant improvement compared to the 84.26% mAP of the original YOLOv5. In addition, we also input the information obtained by the detector into DeepSORT for goat tracking and counting. The average overlap rate of our proposed method is 89.69%, which is significantly higher than the 82.78% of the original combination of YOLOv5 and DeepSORT. In order to avoid double counting as much as possible, goats were counted using the single-line counting based on the results of goat head tracking, which can support practical applications.

show abstract

Statistical Management of Building Fire Equipment Based on Computer Vision and Deep Learning

Zheng,

Tan

2024

Lecture Notes in Operations Research

View full text Add to dashboard Cite

Center clustering network improves piglet counting under occlusion

Cited by 20 publications

References 22 publications

Occlusion-Resistant Instance Segmentation of Piglets in Farrowing Pens Using Center Clustering Network

Occlusion-Resistant Instance Segmentation of Piglets in Farrowing Pens Using Center Clustering Network

An Integrated Goat Head Detection and Automatic Counting Method Based on Deep Learning

Statistical Management of Building Fire Equipment Based on Computer Vision and Deep Learning

Contact Info

Product

Resources

About