206
6 Deep Learning and RFID System Physical Anti-Collision
Fig. 6.4 Flow chart of
multi-label localization
system
improve the image quality through pre-processing and get a potentially sharp image.
As can be seen from Fig. 6.5a, the multi-label image in the horizontal orientation is
set to a black turning disc as the image background, and the label position of the arrow
mark on the tray surface is easily extracted. The acquired image can locate multilabel completely and clearly. In Fig. 6.5b, it is more difficult to process multi-label
image in the complicated background. Therefore, the multi-label image processing
is only for the multi-label image of the complex background in the vertical direction.
6 Deep Learning and RFID System Physical Anti-Collision
Fig. 6.4 Flow chart of
multi-label localization
system
improve the image quality through pre-processing and get a potentially sharp image.
As can be seen from Fig. 6.5a, the multi-label image in the horizontal orientation is
set to a black turning disc as the image background, and the label position of the arrow
mark on the tray surface is easily extracted. The acquired image can locate multilabel completely and clearly. In Fig. 6.5b, it is more difficult to process multi-label
image in the complicated background. Therefore, the multi-label image processing
is only for the multi-label image of the complex background in the vertical direction.