白丝美女被狂躁免费视频网站,500av导航大全精品,yw.193.cnc爆乳尤物未满,97se亚洲综合色区,аⅴ天堂中文在线网官网

Primary user selection for head tracking

專利號(hào)
US10027883B1
公開日期
2018-07-17
申請(qǐng)人
Amazon Technologies, Inc.(US NV Reno)
發(fā)明人
Cheng-Hao Kuo; Jim Oommen Thomas; Tianyang Ma; Stephen Vincent Mangiat; Sisil Sanjeev Mehta; Ambrish Tyagi; Amit Kumar Agrawal; Kah Kuen Fu; Sharadh Ramaswamy
IPC分類
G06K9/00; G06K9/46; G06K9/66; H04N5/225; H04N5/232
技術(shù)領(lǐng)域
face,image,in,tracking,user,bounding,or,device,algorithm,can
地域: NV NV Reno

摘要

Various embodiments enable a primary user to be identified and tracked using stereo association and multiple tracking algorithms. For example, a face detection algorithm can be run on each image captured by a respective camera independently. Stereo association can be performed to match faces between cameras. If the faces are matched and a primary user is determined, a face pair is created and used as the first data point in memory for initializing object tracking. Further, features of a user's face can be extracted and the change in position of these features between images can determine what tracking method will be used for that particular frame.

說(shuō)明書

FIG. 5C illustrates an example process 540 for recovering an undetected face in an image that can be used in accordance with various embodiments. In this example, a first image is acquired from a first camera and a second image is acquired from a second camera 542. Accordingly, a first face and a second face are detected in the first image 544. In this example, only the first face is detected in the second image and the second face is not detected 546. In order to properly match the correct face, the first face and the second face detected in the first image are compared to the first face in the second image 548. Accordingly, the first face from the first image is matched with the first face in the second image 550. Thus, in order to recover the missing second face from the second, a template of the second face is generated from the the second face from the first image 552. In one example, the template can be generated by extracting features of the second face from the first image. Accordingly, the template for the second face is used to analyze and compared to features within the second image using a template matching algorithm 554. In this example, the second face is detected in the second image using the template 556.

權(quán)利要求

1
微信群二維碼
意見(jiàn)反饋