A dataset of grape multimodal object detection and semantic segmentation

The accuracy of grape picking point localization is dependent on grape detection and semantic segmentation network performance. However, in practical application scenarios, the accuracy and segmentation precision of grape targets based on visible light images are susceptible to light variations and complex environments, often performing poorly. Moreover, grapes grow in bunches, and the existing multimodal datasets for apples and pears can hardly meet the recognition needs of bunch-shaped grapes. The construction of visible, depth, and near-infrared multimodal object detection and semantic segmentation datasets of grapes is crucial to exploring better recognition rates and stronger generalization capabilities for grape detection and semantic segmentation models. This dataset, totaling about 39.08 GB, contains high-quality multimodal video stream data of green and purple grapes, including six varieties, under different illumination and obscuration conditions. Additionally, the dataset offers 3954 labeled image samples extracted from the aforementioned multimodal video. By means of rotation, deflation, mis-slicing, panning, and Gaussian blur, the dataset can be augmented for the training implementation of mainstream deep learning models. The dataset can provide valuable basic data resources for multimodal fusion, grape semantic segmentation, and object detection, which have important practical application value for promoting research in the field of agricultural machinery and equipment intelligence.

葡萄采摘点定位的精度，取决于葡萄目标检测（object detection）与语义分割（semantic segmentation）网络的性能。然而在实际应用场景中，基于可见光图像的葡萄目标检测精度与分割精度易受光照变化与复杂环境干扰，往往表现不佳。此外，葡萄呈串状生长，现有针对苹果、梨的多模态数据集（multimodal dataset）难以满足串状葡萄的识别需求。
构建涵盖可见光、深度（depth）与近红外（near-infrared）模态的葡萄目标检测与语义分割数据集，对于研发具备更高识别精度与更强泛化能力的葡萄检测、语义分割模型至关重要。本数据集总容量约39.08 GB，收录了不同光照与遮挡条件下6个品种的绿色与紫色葡萄的高质量多模态视频流数据。此外，本数据集还包含从上述多模态视频中提取的3954张标注图像样本。本数据集可通过旋转、放缩、错切、平移与高斯模糊（Gaussian blur）等方式完成数据增强，以支持主流深度学习模型的训练实施。本数据集可为多模态融合、葡萄语义分割与目标检测任务提供极具价值的基础数据资源，对于推动农业智能装备领域的研究具有重要的实际应用价值。

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于葡萄多模态目标检测和语义分割的数据集，包含约39.08 GB的可见光、深度和近红外多模态视频流数据，涵盖绿色和紫色葡萄的六个品种，并提供了3954个标注图像样本。它旨在解决可见光图像在光照和复杂环境下性能不足的问题，支持数据增强和深度学习模型训练，对多模态融合、葡萄识别及农业机械智能化研究具有重要应用价值。