Abstract: An embodiment of a video-surveillance method comprising: activating a video camera for acquiring, for a surveillance system comprising the aforesaid video camera, a digital video sequence of images of a scene to be surveyed, detecting given events and transmitting to a remote center images of the aforesaid digital video sequence of images acquired according to the result of the aforesaid operation of event detection. An embodiment includes: executing an operation of extraction of key frames on the aforesaid digital video sequence of images for composing a visual storyboard, detecting given events on the basis of data supplied by one or more event sensors for generating an event signal, controlling operating states assumed by the aforesaid key-frame-extraction operation as a function of the values assumed by the aforesaid event signal, and transmitting the aforesaid visual storyboard to the aforesaid remote center.

Abstract: Systems and methods are provided for identifying dark areas in images captured by encapsulated cameras or endoscopes and enhancing the visibility of the dark area using intensity stretch. For capsule or endoscope images, there are often some dark areas caused by inadequate lighting. The poor visibility of detailed features in the dark area may reduce the detection rate of anomaly. The present invention calls for a method to detect the dark area based on contour. The intensities inside and outside the dark area are evaluated and are used to generate intensity transformation to stretch intensity in the dark area while minimizing the impact on the image quality outside the dark area.