Vadzim A. Piatrou,Vladimir N. Linev,Iryna L. Slavashevich,Dmitry V. Pozdnyakov
申请号：
US16150436
公开号：
US10674972B1
申请日:
2018.10.03
申请国别(地区):
US
年份:
2020
代理人:
摘要：
Detecting hidden objects on a human body includes acquiring an incoming X-ray image of the human body passing through a penetrating X-ray scanner； generating additional images based on the incoming image by performing logarithmic or saliency transformations or contrasting of the incoming image； obtaining maps for all objects and known object classes, the maps show which pixels correspond to objects or to background, by passing the incoming and the additional images through a neural network with a deep Segnet-U-Net architecture optimized for overlapping object detection with long skip connections before each downsampling layer of the neural network； using the maps, identifying unknown objects in the incoming image by recognizing all objects/objects of known classes, excluding previously classified objects from the known classes from segmented non-anatomic areas； segmenting the incoming image of the human body into multiple parts； and identifying parts containing objects belonging to both the known and unknown classes.
