翻译：本数据集制作主要依托开源标注工具LabelImg该标注工具可以方便地创建多边形标注框和矩形标注框自动记录框选的位置坐标并将标注信息转化为目标检测领域应用广泛的PASCAL VOC数据集标注格式进行保存减少不同应用指向的检测工作之间的迁移成本提高工作效率标注界面如图2所示。图像标注采取的是单字标注使用LabelImg对图片上对应的甲骨文字符进行框选框选过程中尽量贴紧甲骨文字符如图3所示。每个字

The production of this dataset relies mainly on the open-source annotation tool LabelImg. This tool allows for easy creation of polygon and rectangle annotation boxes, automatically records the selected position coordinates, and saves the annotation information in the widely used PASCAL VOC dataset annotation format for object detection, reducing the transfer cost between detection tasks and improving work efficiency. The annotation interface is shown in Figure 2. The image annotation adopts single-character annotation, using LabelImg to select the corresponding oracle bone character on the image, and trying to fit the oracle bone character as closely as possible during the selection process, as shown in Figure 3. The annotation information for each character is the coordinates (x1, y1, x2, y2) of the selected position, where (x1, y1) is the coordinate of the top-left corner of the rectangle box, and (x2, y2) is the coordinate of the bottom-right corner of the rectangle box