Static surveillance camera.
In this demo can be appreciated, in the left side, the image to be located in the video sequence, while on the right side, it can be seen, in fast-play mode, the video sequence over the searching is performed.
In the last part of this demo, it is shown the identification performed by SIVLI. The yellow markers were put manually, not being part of the system, in order to emphasize how SIVLI recognizes the identified patterns (Please note the Similarity Values presented in the upper left corner). On the other side, there have also been manually marked in blue, those elements that could be a potential source of noise respect to the image to be identified. Note, using the Similarity Value as reference, how SIVLI is tolerant to those sources.
Finally, it is important to show the tolerance to changes in illumination, similar vehicles, people passing, etc. Moreover, the fact the capture by the camera is in practical terms always the same, gives elements to sustain SIVLI’s robustness to minor changes in video sequences with those characteristics.
In order to know more about how SIVLI works it is recommended to visit the SIVLI's Description.
In this demo can be appreciated, in the left side, the image to be located in the video sequence, while on the right side, it can be seen, in fast-play mode, the video sequence over the searching is performed.
Finally, it is important to show the tolerance to changes in illumination, similar vehicles, people passing, etc. Moreover, the fact the capture by the camera is in practical terms always the same, gives elements to sustain SIVLI’s robustness to minor changes in video sequences with those characteristics.
In order to know more about how SIVLI works it is recommended to visit the SIVLI's Description.
Comentarios
Publicar un comentario