Determining region-of-interest of an object using image-based object tracking

A tracking controller, comprising:

• a computer processor;
• and
• memory coupled to the computer processor and storing instructions, when executed, causing the computer processor to:
• capture, using a camera device, a sequence of images of a scene comprising a light source attached to a first element of a plurality of elements comprised in an object;
• detect, based on a pattern of local light change across the sequence of images, the light source in the scene;
• determine a location of the light source in at least one image of the sequence of images;
• generate, based on the location of the light source and a dynamic model of the object, a region-of-interest for analyzing the object;
• and
• generate an analysis result of the object based on the region-of-interest, wherein a pre-determined task is performed based on the analysis result,
• wherein the dynamic model describes mechanical linkages among the plurality of elements,
• wherein the location of the light source in the at least one image corresponds to the location of the first element in the scene, and
• wherein generating the region-of-interest comprises:
• determining, based on the mechanical linkages, a range limitation of a second element of the plurality of elements relative to the first element, and
• determining, based on the location of the light source and the range limitation, a probability map of the location of the second element in the scene.

A system for analyzing an object in a scene, comprising:

• a light source attached to a first element of a plurality of elements comprised in the object;
• a camera device configured to capture a sequence of images of the scene;
• and
• a tracking controller configured to:
• detect, based on a pattern of local light change across the sequence of images, the light source in the scene;
• determine a location of the light source in at least one image of the sequence of images;
• generate, based on the location of the light source and a dynamic model of the object, a region-of-interest for analyzing the object;
• and
• generate an analysis result of the object based on the region-of-interest, wherein a pre-determined task is performed based on the analysis result,
• wherein the dynamic model describes mechanical linkages among the plurality of elements,
• wherein the location of the light source in the at least one image corresponds to the location of the first element in the scene, and
• wherein generating the region-of-interest comprises:
• determining, based on the mechanical linkages, a range limitation of a second element of the plurality of elements relative to the first element, and
• determining, based on the location of the light source and the range limitation, a probability map of the location of the second element in the scene.

A non-transitory computer readable medium storing instructions for analyzing an object in a scene, the instructions, when executed by a computer processor, comprising functionality for:

• capturing, using a camera device, a sequence of images of the scene comprising a light source attached to a first element of a plurality of elements comprised in the object;
• detecting, based on a pattern of local light change across the sequence of images, the light source in the scene;
• determining a location of the light source in at least one image of the sequence of images;
• generating, based on the location of the light source and a dynamic model of the object, a region-of-interest for analyzing the object;
• and
• generating an analysis result of the object based on the region-of-interest, wherein a pre-determined task is performed based on the analysis result,
• wherein the dynamic model comprises nodes representing at least one selected from a group consisting of a head element, a neck element, a pelvis element, a knee element, and an ankle element of the object,
• wherein the pre-determined task is performed by applying at least one selected from a group consisting of a facial expression recognition algorithm and a hand gesture recognition algorithm to the region-of-interest,
• wherein the dynamic model describes mechanical linkages among the plurality of elements,
• wherein the location of the light source in the at least one image corresponds to the location of the first element in the scene, and
• wherein generating the region-of-interest comprises:
• determining, based on the mechanical linkages, a range limitation of a second element of the plurality of elements relative to the first element, and
• determining, based on the location of the light source and the range limitation, a probability map of the location of the second element in the scene.