Corn field multi-type weed real-time identification algorithm based on Kalman filtering and deep learning

• Main Authors: KWON YONGUL, YANG YUNHUAN
• Format: Patent
• Language: Chinese; English
• Published: 30.05.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a corn field multi-type weed real-time identification algorithm based on Kalman filtering and deep learning, and belongs to the field of deep learning and precision agriculture. In a seedling grass identification process, Kalman filtering and an IOU matching technology are fused to form a stable tracking algorithm, and a color segmentation technology and morphological analysis processing are combined to realize identification of weeds with multiple types and multiple feature exposure degrees. The recognition rate of corn seedlings and weeds through an artificial intelligence algorithm is greatly improved, and a foundation is laid for precision agriculture. 本发明公开了一种基于卡尔曼滤波和深度学习的玉米田地多种类杂草实时识别算法，属于深度学习和精准农业领域。在对苗草识别的过程中，融合卡尔曼滤波和IOU匹配技术构成稳定追踪算法，再结合颜色分割技术和形态学分析处理，实现多种类、多特征暴露程度杂草的识别。大幅提高通过人工智能算法对玉米苗和杂草的识别率，对精准农业打下基础。