猪只图片数据集

数据采集：从上百个养猪场的现场摄像头画面抽帧，以及外勤人员手动拍摄采集的图片和视频数据集；
数据处理：使用人工筛选标注，然后通过贵州东彩独有的生猪识别模型、生猪视频点数模型训练100次以上；
数据应用：可用于猪只目标识别、猪只行为分析、猪只图片/视频点数。

Data Collection: This dataset is compiled by extracting frames from on-site camera footage of hundreds of pig farms, as well as images and videos manually captured by field personnel.
Data Processing: First, manual screening and annotation are conducted on the collected data. Subsequently, the dataset is trained for over 100 epochs using Guizhou Dongcai's proprietary swine recognition model and pig video counting model.
Data Applications: This dataset can be applied to pig target recognition, pig behavior analysis, and pig counting in both images and videos.

搜集汇总

背景与挑战

背景概述

猪只图片数据集由贵州东彩科技供应链科技有限公司申请，数据规模为12181.9MB，更新周期为月。该数据集主要用于现代化猪只养殖管理、金融和保险行业的监管与价值评估，数据来源于公开收集，采集方式包括摄像头画面抽帧和外勤人员手动拍摄，经过人工筛选标注和模型训练。