基于改进Mask2Former的羊群图像实例分割研究

仇 旋旋
重庆三峡学院

摘要


针对畜牧场景中羊只个体频繁交叠、背景环境干扰剧烈等导致的分割难题设计名为DM-Former的增强型实
例分割网络。该模型以Mask2Former为基础框架,通过在像素解码器层级引入多尺度空洞融合(MDFA)单元,大幅
提升对重叠目标的全局依赖建模能力;同时,结合中值增强空间通道注意力(MECS)单元,实现对牧场复杂纹理噪
声的高效抑制。实验数据表明,在自建的羊群图像数据集中,DM-Former的平均精度(AP)达到了82.69%。与现有
基准模型相比,DM-Former方法在保持分割完整度的同时,优化边缘的精细化程度,为实现精准畜牧业的个体监测
提供核心技术依据。

关键词


改进Mask2Former;实例分割;精准畜牧业;注意力机制;空洞卷积

全文:

PDF


参考


[1]He K, Gkioxari G, Dollár P, et al. Mask r-cnn[C]//

Proceedings of the IEEE international conference on computer

vision. 2017: 2961-2969.

[2]Bhujel A, Wang Y, Lu Y, et al. Public computer vision

datasets for precision livestock farming: A systematic survey[J].

arXiv preprint arXiv:2406.10628, 2024.

[3]Cheng B, Misra I, Schwing A G, et al. Maskedattention mask transformer for universal image segmentation[C]//

Proceedings of the IEEE/CVF conference on computer vision

and pattern recognition. 2022: 1290-1299.

[4]Liu Z, Lin Y, Cao Y, et al. Swin transformer:

Hierarchical vision transformer using shifted windows[C]//

Proceedings of the IEEE/CVF international conference on

computer vision. 2021: 10012-10022.

[5]Ghiasi G, Cui Y, Srinivas A, et al. Simple copy-paste is a

strong data augmentation method for instance segmentation[C]//

Proceedings of the IEEE/CVF conference on computer vision

and pattern recognition. 2021: 2918-2928.

[6]Ghiasi G, Cui Y, Srinivas A, et al. Simple copy-paste is a

strong data augmentation method for instance segmentation[C]//

Proceedings of the IEEE/CVF conference on computer vision

and pattern recognition. 2021: 2918-2928.


Refbacks

  • 当前没有refback。