METHOD AND DEVICE FOR DE-IDENTIFYING PERSONAL INFORMATION CONSECUTIVELY IN DEEP LEARNING-BASED VIDEO

摘要

The present invention relates to a method and device for de-identifying personal information consecutively in a deep learning-based video. A method for de-identifying personal information consecutively in a deep learning-based video by a device for de-identifying personal information consecutively in a deep learning-based video according to an embodiment of the present invention comprises the steps of: detecting a first detection area of a particular object and masking the first detection area; detecting a second detection area of the particular object and masking the second detection area; checking whether or not the first detection area and second detection area are correlated; if the first detection area and second detection area are correlated, checking a non-detection frame between a first detection area frame comprising the first detection area and a second detection area frame comprising the second detection area; and calculating an estimation area of the particular object from the non-detection frame, and masking the estimation area.