An image acquisition device compresses a media signal representative of a scene based on a sensing matrix that is a determined by a sensing matrix template and a set of template parameters. The image acquisition device provides the compressed media signal to a receiver and selectively provides a specification of a subset of the set of template parameters to the receiver. The receiver extracts one or more scene descriptors representative of one or more portions of the scene from the compressed media signal using the sensing matrix template without knowledge of the template parameters that are not included in the subset. The template parameters that are not included in the subset are not received by the receiver.

A device may include one or more processors, an image signal processor, and system memory. The one or more processors are configured to receive a request to accommodate a preview stream and a snapshot stream and output a request to increase a bandwidth between the image signal processor and the system memory to accommodate the preview stream in response to receiving the request to accommodate the preview stream and the snapshot stream. The request to increase the bandwidth to accommodate the preview stream refrains from accommodating the snapshot stream. Responsive to receiving a request to generate an image, for storage at the system memory, that corresponds to an image being output by the preview stream, the one or more processors are further configured to output a request to further increase the bandwidth between the image signal processor and the system memory to accommodate the snapshot stream.

A system for event recording includes an interface and a processor. The interface is configured to receive video data and sensor data. The processor is configured to determine a risk level based at least in part on the sensor data; determine a storage quality based at least in part on the risk level; determine a processed video data by processing the video data according to the storage quality; and store the processed video data.

An image capturing apparatus comprises: an image sensor configured to output image data captured at any of two or more sensitivities; and a memory storing instructions which cause the apparatus to function as: a control unit configured to control a compression rate for compressing the outputted image data; a compression unit configured to compress the outputted image data at the compression rate; a decompression unit configured to decompress the compressed image data; and an application unit configured to apply a gain amount to the decompressed image data based on the two or more sensitivities, wherein the control unit controls the compression rate so that a compression rate in a case where a first gain amount is applied to is less than a compression rate in a case where a second gain amount is applied to.

An image capturing apparatus comprises: an image sensor configured to output image data captured at any of two or more sensitivities; and a memory storing instructions which cause the apparatus to function as: a control unit configured to control a compression rate for compressing the outputted image data; a compression unit configured to compress the outputted image data at the compression rate; a decompression unit configured to decompress the compressed image data; and an application unit configured to apply a gain amount to the decompressed image data based on the two or more sensitivities, wherein the control unit controls the compression rate so that a compression rate in a case where a first gain amount is applied to is less than a compression rate in a case where a second gain amount is applied to.

The present disclosure provides systems and methods that are configured for feature extraction or object recognition using compressive measurements that represent a compressed image of a scene. In various aspects, a compressive sensing matrix is constructed and used to acquire the compressive measurements, such that in the extraction phase, the compressive measurements can be processed to detect feature points and determine their feature vectors in the scene without using a pixel representation of the scene. The determined feature vectors are used to detect objects based on comparison with one or more predetermined feature vectors.

The present disclosure provides systems and methods that are configured for feature extraction or object recognition using compressive measurements that represent a compressed image of a scene. In various aspects, a compressive sensing matrix is constructed and used to acquire the compressive measurements, such that in the extraction phase, the compressive measurements can be processed to detect feature points and determine their feature vectors in the scene without using a pixel representation of the scene. The determined feature vectors are used to detect objects based on comparison with one or more predetermined feature vectors.

Techniques and devices for acquiring and compressing timelapse video are described. The techniques are adaptive, in that the acquisition frame rate is adapted and captured images are periodically deleted during filming. To implement the method, a user need not know ahead of time how long the source video will be acquired. Regardless of the acquisition time, the resulting video is automatically edited to provide a timelapse clip of a predefined length or of a length within a predefined range.

Examples are disclosed for video analytics of captured video content. In some examples, information may be received from a host processing system for a camera to capture video content. The camera may be a surveillance camera or a camera located with a display device. Video analytics may be performed on the captured video and the captured video content may be encoded. Data associated with the video analytics may then be sent to the host processing system. In some examples, the data as well as encoded captured video content or streaming video may be sent via communication channels included in an interconnect. Other examples are described and claimed.

Examples are disclosed for video analytics of captured video content. In some examples, information may be received from a host processing system for a camera to capture video content. The camera may be a surveillance camera or a camera located with a display device. Video analytics may be performed on the captured video and the captured video content may be encoded. Data associated with the video analytics may then be sent to the host processing system. In some examples, the data as well as encoded captured video content or streaming video may be sent via communication channels included in an interconnect. Other examples are described and claimed.