English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Computer vision enables the development of new approaches to monitor thebehavior, health, and welfare of animals. Instance segmentation is ahigh-precision method in computer vision for detecting individual animals ofinterest. This method can be used for in-depth analysis of animals, such asexamining their subtle interactive behaviors, from videos and images. However,existing deep-learning-based instance segmentation methods have been mostlydeveloped based on public datasets, which largely omit heavy occlusionproblems; therefore, these methods have limitations in real-world applicationsinvolving object occlusions, such as farrowing pen systems used on pig farms inwhich the farrowing crates often impede the sow and piglets. In this paper, weadapt a Center Clustering Network originally designed for counting to achieveinstance segmentation, dubbed as CClusnet-Inseg. Specifically, CClusnet-Inseguses each pixel to predict object centers and trace these centers to form masksbased on clustering results, which consists of a network for segmentation andcenter offset vector map, Density-Based Spatial Clustering of Applications withNoise (DBSCAN) algorithm, Centers-to-Mask (C2M), and Remain-Centers-to-Mask(RC2M) algorithms. In all, 4,600 images were extracted from six videoscollected from three closed and three half-open farrowing crates to train andvalidate our method. CClusnet-Inseg achieves a mean average precision (mAP) of84.1 and outperforms all other methods compared in this study. We conductcomprehensive ablation studies to demonstrate the advantages and effectivenessof core modules of our method. In addition, we apply CClusnet-Inseg tomulti-object tracking for animal monitoring, and the predicted object centerthat is a conjunct output could serve as an occlusion-resistant representationof the location of an object.

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