Camera parameter control using face vectors for portal

Abstract

A method of performing adjustment of a convergence speed for an imaging parameter is based on detecting a motion of a selected face in a sequence of image frames. When the detected motion meets predefined motion criteria, a motion vector corresponding to the characterized motion of the face is computed. A value for a convergence adjustment factor for adjusting a convergence speed of an imaging parameter of the camera is determined based on the computed motion vector. The convergence speed of the imaging parameter of the camera is adjusted based on the determined value of the convergence adjustment factor.

What is claimed is: 
     
         1 . A method comprising:
 receiving a sequence of image frames of video data of a scene from a camera on a client device;   detecting motion data characterizing motion of an object in the scene from the received sequence of image frames;   determining that the detected motion data meets one or more predefined motion criteria;   responsive to determining that the detected motion data meets the one or more predefined motion criteria, computing information describing the characterized motion of the object;   determining a value of a convergence adjustment factor based on the computed motion vector for adjusting a convergence speed of an imaging parameter of the camera; and   adjusting the convergence speed of the imaging parameter of the camera based on the determined value of the convergence adjustment factor.   
     
     
         2 . The method of  claim 1 , wherein detecting motion data characterizing motion of the object in the scene from the received sequence of image frames comprises:
 receiving a list of one or more detected objects for a current image frame;   selecting an object from the received list of one or more detected objects for the current image frame;   determining that the selected object for the current image frame is the same as a selected object for a previous image frame; and   calculating a value associated with a motion of the selected object for the current image frame between the current image frame and the previous image frame.   
     
     
         3 . The method of  claim 2 , wherein selecting the object from the received list of one or more detected objects further comprises at least one of:
 determining that the selected object is a largest object from the received list of one or more detected objects based on parameter values associated with the detected obj ects; and   determining that the selected object is located closest to the center of the current image frame in at least one of the current image frame and the previous image frame from the received list of one or more detected object based the parameter values associated with the detected objects.   
     
     
         4 . The method of  claim 2 , wherein calculating the value associated with the distance moved by the selected object between the current image frame and the previous image frame comprises:
 calculating an area of overlap of bounding boxes for a location of the selected object in each of the current image frame and the previous image frame relative to an area of union of the bounding boxes for the location of the selected object in each of the current image frame and the previous image frame.   
     
     
         5 . The method of  claim 2 , wherein determining that the detected motion data meets the one or more predefined motion criteria comprises:
 computing a proportion of prior frames in which the value for the motion of the selected object between the current image frame and the previous image frame exceeds a predefined threshold motion value; and   responsive to determining that the computed proportion of prior frames exceeds a predefined threshold number of frames, determining that the motion data value meets the one or more predefined motion criteria.   
     
     
         6 . The method of  claim 1 , wherein computing information describing the characterized motion of the object comprises:
 establishing a search window that comprises a minimum rectangle that includes bounding boxes corresponding to the object in a pair of consecutive frames of the received sequence of video frames; and   computing a motion vector based on minimizing a cost function using block matching over the established search window.   
     
     
         7 . The method of  claim 6 , wherein the cost function comprises minimizing a sum of differences of a function based on red-green-blue pixel values. 
     
     
         8 . The method of  claim 6 , wherein computing the motion vector based on minimizing the cost function using block matching over the established search window comprises computing the motion vector that generates the minimum cost function when performing block matching over the established search window. 
     
     
         9 . The method of  claim 1 , wherein determining the value of the convergence adjustment factor based on the information describing the characterized motion of the object for adjusting a convergence speed of an imaging parameter of the camera comprises applying a model that determines the value of the convergence adjustment factor for the computed information describing the characterized motion of the object. 
     
     
         10 . The method of  claim 9 , wherein applying the model that determines the value of the convergence adjustment factor for the computed information describing the characterized motion of the object vector comprises:
 in response to determining that the information describing the characterized motion of the object is above a predefined first threshold, linearly decreasing the value of the convergence adjustment factor based on the computed information describing the characterized motion of the object; and   in response to determining that the computed information describing the characterized motion of the object is above a predefined second threshold, the predefined second threshold being greater than the predefined first threshold, establishing the value of the convergence adjustment factor as zero.   
     
     
         11 . A non-transitory computer-readable medium comprising computer program instructions that, when executed by a computer processor of an online system, cause the processor to perform steps comprising:
 receive a sequence of image frames of video data of a scene from a camera on a client device;   detect motion data characterizing motion of an object in the scene from the received sequence of image frames;   determine that the detected motion data meets one or more predefined motion criteria;   responsive to determining that the detected motion data meets the one or more predefined motion criteria, compute information describing the characterized motion of the object;   determine a value of a convergence adjustment factor based on the computed motion vector for adjusting a convergence speed of an imaging parameter of the camera; and   adjust the convergence speed of the imaging parameter of the camera based on the determined value of the convergence adjustment factor.   
     
     
         12 . The non-transitory computer readable medium of  claim 11 , wherein detect motion data characterizing motion of the object in the scene from the received sequence of image frames comprises:
 receive a list of one or more detected objects for a current image frame;   select an object from the received list of one or more detected objects for the current image frame;   determine that the selected object for the current image frame is the same as a selected object for a previous image frame; and   calculate a value associated with a motion of the selected object for the current image frame between the current image frame and the previous image frame.   
     
     
         13 . The non-transitory computer readable medium of  claim 12 , wherein select the object from the received list of one or more detected objects further comprises at least one of:
 determine that the selected object is a largest object from the received list of one or more detected objects based on parameter values associated with the detected obj ects; and   determine that the selected object is located closest to the center of the current image frame in at least one of the current image frame and the previous image frame from the received list of one or more detected object based the parameter values associated with the detected objects.   
     
     
         14 . The non-transitory computer readable medium of  claim 12 , wherein calculate the value associated with the distance moved by the selected object between the current image frame and the previous image frame comprises:
 calculate an area of overlap of bounding boxes for a location of the selected object in each of the current image frame and the previous image frame relative to an area of union of the bounding boxes for the location of the selected object in each of the current image frame and the previous image frame.   
     
     
         15 . The non-transitory computer readable medium of  claim 12 , wherein determine that the detected motion data meets the one or more predefined motion criteria comprises:
 compute a proportion of prior frames in which the value for the motion of the selected object between the current image frame and the previous image frame exceeds a predefined threshold motion value; and   responsive to determining that the computed proportion of prior frames exceeds a predefined threshold number of frames, determine that the motion data value meets the one or more predefined motion criteria.   
     
     
         16 . The non-transitory computer readable medium of  claim 11 , wherein compute information describing the characterized motion of the object comprises:
 establish a search window that comprises a minimum rectangle that includes bounding boxes corresponding to the object in a pair of consecutive frames of the received sequence of video frames; and   compute a motion vector based on minimizing a cost function using block matching over the established search window.   
     
     
         17 . The non-transitory computer readable medium of  claim 16 , wherein the cost function comprises minimizing a sum of differences of a function based on red-green-blue pixel values. 
     
     
         18 . The non-transitory computer readable medium of  claim 16 , wherein compute the motion vector based on minimizing the cost function using block matching over the established search window comprises computing the motion vector that generates the minimum cost function when performing block matching over the established search window. 
     
     
         19 . The non-transitory computer readable medium of  claim 11 , wherein determine the value of the convergence adjustment factor based on the information describing the characterized motion of the object for adjusting a convergence speed of an imaging parameter of the camera comprises applying a model that determines the value of the convergence adjustment factor for the computed information describing the characterized motion of the object. 
     
     
         20 . The non-transitory computer readable medium of  claim 19 , wherein applying the model that determines the value of the convergence adjustment factor for the computed information describing the characterized motion of the object vector comprises:
 in response to determining that the information describing the characterized motion of the object is above a predefined first threshold, linearly decreasing the value of the convergence adjustment factor based on the computed information describing the characterized motion of the object; and   in response to determining that the computed information describing the characterized motion of the object is above a predefined second threshold, the predefined second threshold being greater than the predefined first threshold, establishing the value of the convergence adjustment factor as zero.