Data annotation target auxiliary positioning method, data annotation system, equipment and storage medium

The invention discloses a data annotation target auxiliary positioning method, a data annotation system, equipment and a storage medium, and the method comprises the following steps: selecting and training a target detection model for auxiliary positioning according to the type of an annotated target image data set; using the trained target detection model to process a to-be-labeled image to obtain a to-be-labeled target sequence, including the position, size and confidence of a to-be-labeled target; according to the selected to-be-labeled targets in the to-be-labeled target sequence, the selected to-be-labeled targets are translated and zoomed into a preset labeling work view frame. According to the method, the labeling target can be automatically translated and zoomed to a proper working view, and the labeling efficiency is improved. 本发明公开了一种数据标注目标辅助定位方法、数据标注系统、设备及存储介质，包括步骤如下：根据已标注的目标图像数据集的类型，选择并训练用于辅助定位的目标检测模型；使用完成训练的目标检测模型对待标注的图像进行处理，得到待标注目标序列，包括待标注目标的位置、大小、置信度；根据选择待标注目标序列中的待标注目标，将选择的待标注目标平移、缩放至预先设置的标注工作视