一种课堂面部数据集的构建方法及系统

• Format: Patent
• Language: Chinese
• Published: 15.12.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

本发明一种课堂面部数据集的构建方法及系统，属于图像处理技术领域；所述构建方法包括构建能表明受试者唯一身份的身份标签列表，并确定受试者的基准面部图像；逐帧截取课堂监控视频画面，将同一受试者图像存储到以受试者身份标签命名的文件夹中；使用OpenCV、SSD等五个面部检测算法检测面部图像并保存；对同一文件夹中的图像做数据增强；使用训练好的DeepFace模型提取面部特征向量；基于面部特征向量，使用Density Peak Clustering聚类算法去除低质量的数据样本；对过滤后的数据进行清洗，得到所述课堂面部数据集。本发明提出的课堂面部数据集构建方法能够得到高质量的大规模数据集，同时能够有效提升面部识别模型的识别精度。 The invention discloses a construction method and system of a classroom face data set, and belongs to the technical field of image processing. The construction method comprises the following steps: constructing an identity tag list capable of indicating the unique identity of a subject, and determining a reference facial image of the subject; intercepting classroom monitoring video pictures frame by frame, and storing the same subject image in a folder named by a subject identity tag; five face detection algorithms such as OpenCV and SSD are used for detecting and storing face images; performing data enhancement on the images in the same folder; extracting a facial feature vector by using the trained De