Multi-stream hybrid architecture based on cross-level fusion strategy for fine-grained crop species recognition in precision agriculture
Multi-stream hybrid architecture based on cross-level fusion strategy for fine-grained crop species recognition in precision agriculture
复制标题
DOI：
10.1016/j.compag.2021.106134
复制
发表时间：
2021-04-07
影响因子：
8.3
通讯作者：
Lin, Seng
中科院分区：
文献类型：
--
作者：
Kong, Jianlei;Wang, Hongxing;Lin, Seng
精准农业旨在优化作物生产流程，尽可能更高效、更合理地管理可持续的供应链实践。近年来，深度学习、物联网等各种先进技术在现实农业条件下取得了显著的智能进步。然而，农作物物种识别问题可以看作是细粒度视觉分类(FGVC)问题，具有较低的类间差异和较高的类内差异，这比依赖传统的深度神经网络(DNN)的普通基本类别分类更具挑战性。针对实际农田场景中不同作物种类的分类问题，提出了一种基于细粒度的视觉识别模型MCF-Net。提出的MCF-Net网络由跨级部分网络(CSPNet)作为骨干模块、三个并行子网络和跨级融合模块组成。通过利用海量细粒度信息的多数据流混合结构，MCF-Net在区分类间差异和容忍类内差异方面具有较好的表示能力。此外，通过跨层融合策略对MCF-Net的端到端实现进行了优化，以准确识别不同的作物类别。在现场CropDeepv2数据集上的几个实验表明，我们的方法比最先进的方法更有利。MCF-Net的识别率和F1-Score分别达到了90.6%和0.962，均优于对比模型，表明该方法具有较好的识别精度和模型稳定性。此外，MCF-Net的总体参数仅为807M字节，在模型性能和复杂度之间取得了很好的平衡。物联网平台在精准农业实践中的实施是可以接受和适合的。
Precision farming aims to optimizing the crop production process and managing sustainable supply chain practices as more efficient and reasonable as possible. Recently, various advanced technologies, such as deeplearning and internet of things (IoT), have achieved remarkable intelligence progress in realistic agricultural conditions. However, crops species recognition can be considered as fine-grained visual classification (FGVC) problem, suffering the low inter-class discrepancy and high intra-class variances from the subordinate categories, which is more challenging than common basic-level category classification depended on traditional deep neural networks (DNNs). This paper presents a fine-grained visual recognition model named as MCF-Net to classifying different crop species in practical farmland scenes. Proposed MCF-Net is consisted of cross stage partial network (CSPNet) as backbone module, three parallel sub-networks, and cross-level fusion module. With multi-stream hybrid architecture utilizing massive fine-granulometric information, MCF-Net obtains preferable representation ability for distinguishing interclass discrepancy and tolerating intra-class variances. In addition, the end-toend implementation of MCF-Net is optimized by cross-level fusion strategy to accurately identify different crop categories. Several experiments on in-field CropDeepv2 datasets demonstrate that our method favorably against the state-of-the-art methods. The recognition accuracy and F1-score of MCF-Net achieves very competitive results up to 90.6% and 0.962 separately, both of which outperforming contrasted models indicate better recognition accuracy and model stability of our method. Moreover, the overall parameters of MCF-Net are only 807 MByte with achieving a good balance between model's performance and complexity. It is acceptable and suitable to the implementation of IoT platforms in precision agricultural practices.