Abstract: Systems and methods for image analysis and recognition are disclosed, in particular the methods for interest point description. An interest point and its surrounding area is broken into subareas, a frequency domain description of each area is created by applying discrete Fourier transform (DFT). Frequency domain features are than coded bitwise by comparing them to predefined thresholds. Subsequently, the present invention provides alternative or improved methods and data structures for interest point description that may reduce memory consumption and allow fast bitwise matching.

Abstract: Systems and methods are disclosed for correlating sensor inputs with content stream intervals and selectively requesting and transmitting content streams. In one implementation, a processing device receives a content stream from a media capture device. The processing device correlates one or more chronological intervals of the content stream with one or more first inputs, the one or more first inputs originating from one or more sensors. The processing device processes the one or more first inputs to identify an incidence of one or more phenomena. Based on an identification of the incidence of the one or more phenomena, the processing device selects at least one of the one or more chronological intervals of the content stream that correspond to the one or more phenomena.