Night small target fruit detection method and system fusing global fine-grained information

• Main Authors: JIA WEIKUAN, SUN MEILI, ZHENG YUANJIE, WEI JINMENG, XU LIANCHENG
• Format: Patent
• Language: Chinese; English
• Published: 27.05.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention provides a night small target fruit detection method fusing global fine-grained information, and the method is characterized in that the method comprises the steps: obtaining a to-be-detected fruit image; performing feature extraction on the fruit image based on a pre-trained residual network model; wherein the residual network model comprises a plurality of convolution blocks and a focus bottleneck Transform structure, and coarse-grained information and local fine-grained information are extracted from a global region through the focus bottleneck Transform structure; taking the features output by each layer of the residual network model as the input of each layer of a feature pyramid network model, and outputting feature information under different scales; and inputting the feature information into a pre-trained target detector, and outputting a target detection result. 本公开提供了一种融合全局细粒度信息的夜间小目标果实检测方法，其特征在于，包括：获取待检测的果实图像；基于预先训练的残差网络模型，对所述果实图像进行特征提取；其中，所述残差网络模型包括若干卷积块及焦点瓶颈Transformer结构，通过所述焦点瓶颈Tra