Effect of image restoration on acquisition of moving
objects from thermal video sequences degraded by the atmosphere
By: Oren Haik and Yitzhak Yitzhaky
Abstract
Remotely sensed videos captured by
high-resolution imagers, are likely to be degraded by the atmosphere. The
degradation sources that include turbulence and aerosols cause mainly image
blur in still images. In video sequences however, spatiotemporal-varying distortions
caused by turbulence become also meaningful. These atmospheric degradations
reduce image quality, and therefore, the ability of target acquisition by the
observers. The effects of image quality and image restoration (de-blurring) on
target acquisition in still images were examined previously in several studies.
Nevertheless, results obtained in static situations may not be appropriate for
dynamic situations (with moving targets), which are frequently more realistic.
This work examines the effects of image restoration on the ability of
observers to acquire moving objects (such as humans and vehicles) in video
sequences. This is done through perception experiments that compare
acquisition probabilities in both restored and non-restored video sequences
captured by a remote-sensing thermal imaging system. Results show that image
restoration may significantly improve the acquisition probability. These
results correspond to the static case. However, unlike the static case,
considerably smaller differences were obtained here between the probabilities
of target detection and target recognition.
Recorded versus
restored videos:
Contrast and brightness in the media player may be adjusted to better view the videos.
Experiment 1: Detection and recognition of moving objects
in the six video clips shown below
-
Video_1 (Construction workers)
Video_2 (Walking person, two dogs)
Video_3 (Two trucks, car)
Video_4 (Walking person, dog, bird)
Video_5 (Forklift, truck)
Video_6 (Bicycle rider)
Experiment 2: Determination
whether a moving person carries a thin pole (a rifle-resembled object) in the
three video clips
shown below -
Video_7 (Walking with a pole)
Video_8 (Walking without a pole)
Video_9 (Walking with a pole)