Systems and methods for maintaining multiple objects within a camera field-of-view

A non-transitory computer-readable medium that stores instructions that can be executed by a processor for controlling a camera to simultaneously maintain multiple objects within a field of view of the camera, comprising:

• logic configured to identify a plurality of object tracking constraint functions to be enforced with respect to the multiple objects that are simultaneously maintained within the field of view, each of the plurality of object tracking constraint functions relating to an attribute of the multiple objects determined from image data captured by the camera, wherein one object tracking constraint function of the plurality of object tracking constraint functions relates to a location of a mean position of the multiple objects within the field of view;
• logic configured to identify a priority ranking for the plurality of object tracking constraint functions such that more important object tracking constraint functions have a higher priority ranking than less important object tracking constraint functions, which have a lower priority ranking;
• logic configured to determine each attribute from the image data captured by the camera, including the mean position of the multiple objects;
• and
• logic configured to determine a set of solutions that satisfy the plurality of object tracking constraint functions relative to an order of their priority ranking such that solutions that satisfy lower priority ranking object tracking constraint functions are only considered viable if they also satisfy any higher priority ranking object tracking constraint functions, each solution of the set of solutions providing an indication as to how to control the camera to simultaneously maintain the multiple objects within the field of view of the camera.

A method for controlling a camera to simultaneously maintain multiple objects within a field of view of the camera, the method comprising:

• developing an underdetermined task function for each of multiple object tracking constraint functions to be enforced with respect to the multiple objects that are simultaneously maintained within the field of view, each of the multiple object tracking constraint functions relating to an attribute of the multiple objects determined from image data captured by the camera, a first object tracking constraint function of the multiple object tracking constraint functions relating to a location of a mean position of the multiple objects that are simultaneously maintained within the field of view and a second object tracking constraint function of the multiple object tracking constraint functions relating to a maximum variance from the mean position of the multiple objects that is permitted for any object of the multiple objects;
• determining each attribute from the image data captured by the camera;
• assigning a priority rank to each of the multiple object tracking constraint functions such that more important object tracking constraint functions have a higher priority rank than less important object tracking constraint functions, the first object tracking constraint function being assigned a higher priority rank than the second object tracking constraint function;
• developing a Jacobian matrix for each of the underdetermined task functions and therefore each of the multiple object tracking constraint functions;
• determining null spaces for each of the Jacobian matrices relative to an order of the priority ranks assigned to their associated object tracking constraint functions such that the null spaces for each of the associated object tracking constraint functions are within the null spaces of any of the multiple object tracking constraint functions having a higher priority rank, each of the null spaces identifying a camera configuration in which the multiple objects will be simultaneously maintained within the field of view;
• and
• outputting camera configuration information to a camera controller that controls the camera configuration, the camera configuration information simultaneously maintaining the multiple objects within the field of view of the camera.