Video camera and method for analyzing a video stream

A method for analyzing a video stream, the method comprising:

• capturing the video stream, wherein the video stream comprises image data from a plurality of image frames;
• storing the image data from the plurality of image frames in an image buffer;
• receiving first sensor data from a sensor, wherein the sensor is at least one of a RADAR sensor, a LIDAR sensor a PIR sensor and/or a microphone,
• performing a first evaluation on a subset of the plurality of image frames by temporally correlating the first sensor data with the subset of the image frames in the video stream wherein the first evaluation generates a corresponding first evaluation result and has an evaluation time larger than 1/fps for the video stream;
• for each image frame in the subset, marking the image frame as being under evaluation to form marked image frames;
• deleting, from the image buffer, one or more image frames among the plurality of images frames that temporally proceed the marked image frames;
• selectively performing a second evaluation on the image data of each of the image frames in the subset based on the corresponding first evaluation result, wherein the second evaluation generates a corresponding second evaluation result;
• outputting the corresponding second evaluation result;
• and
• deleting the marked image frames from the image buffer in response to the outputting.

A non-transitory computer-readable storage medium having stored thereon instructions for analyzing a video stream, the instruction when executed by a processor cause the processor to perform a method comprising:

• capturing the video stream, wherein the video stream comprises image data from a plurality of image frames;
• storing the image data from the plurality of image frames in an image buffer;
• receiving first sensor data from a sensor, wherein the sensor is at least one of a RADAR sensor, a LIDAR sensor a PIR sensor and/or a microphone,
• performing a first evaluation on a subset of the plurality of image frames by temporally correlating the first sensor data with the subset of the image frames in the video stream wherein the first evaluation generates a corresponding first evaluation result and has an evaluation time larger than 1/fps for the video stream;
• for each image frame in the subset, marking the image frame as being under evaluation to form marked image frames;
• deleting, from the image buffer, one or more image frames among the plurality of images frames that temporally proceed the marked image frames;
• selectively performing a second evaluation on the image data of each of the image frames in the subset based on the corresponding first evaluation result, wherein the second evaluation generates a corresponding second evaluation result;
• outputting the corresponding second evaluation result;
• and
• deleting the marked image frames from the image buffer in response to the outputting.