Object Detection Techniques
AssigneeMicrosoft Corporation
InventorsAndrew Sherman Nielsen, Yongjun Wu
Life of this patent
Estimating the term…
Abstract
Object detection techniques are described. In one or more implementations, a plurality of images are received by a computing device. The plurality of images are analyzed by the computing device to detect whether the images include, respectively, a depiction of an object. If an object is found in a first image, the locations, angles and scales for object detection can be further restricted in a second one. If an object is not found in a first one of the image, different portions of a second one of the images are analyzed for object detection.
Claims20
A method comprising:
- receiving a plurality of images by a computing device;
- analyzing the plurality of images by the computing device to detect whether the images include, respectively, a depiction of an object, the analyzing performed such that one or more portions used to detect the object in a first said image is different than one or more portions used to detect the object in a second said image.
A computing device comprising:
- a processing system including a plurality of processor cores, each of the plurality of processor cores is configured to execute one or more threads such that the processing system is configured to execute a plurality of the threads in parallel;
- and
- memory configured to store one or more instructions that are executable to cause the processor system to perform operations comprising analyzing one or more images to detect an object in which performance of the analysis is separated for execution on the plurality of processor cores such that the processor cores utilize different scales or angles, one to another, to perform the analysis to detect the object.
A method comprising:
- determining by a computing device as to whether a human face has previously been detected in one or more frames of images captured by one or more cameras;
- responsive to a determination that the human face has been detected, utilizing one or more characteristics of the previous detection of the human face to detect the human face in one or more subsequent frames of images;
- and
- responsive to a determination that the human face has not been detected, employing a plurality of processor cores of the computing device to analyze the one or more subsequent frames to perform the detection in which performance of the analysis is separated for execution on the plurality of processor cores such that the processor cores utilize different scales or angles, one to another, to perform the analysis to detect the human face.
Dependencies are inferred from the claim text.