-
公开(公告)号：US20150193941A1
公开(公告)日：2015-07-09
申请号：US14149839
申请日：2014-01-08
Inventor： Yan Wang
CPC classification number: G06T7/0085 , G06F3/038 , G06F3/042 , G06F3/0425 , G06T7/13 , G06T7/174
Abstract: A method of determining an edge of an object on a digital image sequence comprising the step of determining a first gradient direction profile of a first image in the digital image sequence; determining a second gradient direction profile of a second image in the digital image sequence; computing a differential profile based on the first gradient direction profile and the second gradient direction profile; and determining the edge of the object based on the differential profile wherein the differential profile registers gradient magnitudes of the second gradient direction profile and angular differences between the first gradient direction profile and the second gradient direction profile. A system thereof is also disclosed.
Abstract translation: 一种确定数字图像序列上的对象边缘的方法，包括确定数字图像序列中的第一图像的第一梯度方向轮廓的步骤; 确定所述数字图像序列中的第二图像的第二梯度方向轮廓; 基于所述第一梯度方向轮廓和所述第二梯度方向轮廓来计算差分轮廓; 以及基于所述差分轮廓来确定所述对象的边缘，其中所述差分轮廓记录所述第二梯度方向轮廓的梯度大小和所述第一梯度方向轮廓与所述第二梯度方向轮廓之间的角度差。 还公开了其系统。
-
公开(公告)号：US10685428B2
公开(公告)日：2020-06-16
申请号：US16185860
申请日：2018-11-09
Inventor： Yan Wang , Hailiang Li , Yang Liu , Man Yau Chiu , Zhi Bin Lei
Abstract: Methods and systems which provide super-resolution synthesis based on weighted results from a random forest classifier are described. Embodiments apply a trained random forest classifier to low-resolution patches generated from the low-resolution input image to classify the low-resolution input patches. As each low-resolution patch is fed into the random forest classifier, each decision tree in the random forest classifier “votes” for a particular class for each of the low-resolution patches. Each class is associated with a projection matrix. The projection matrices output by the decision trees are combined by a weighted average to calculate an overall projection matrix corresponding to the random forest classifier output, which is used to calculate a high-resolution patch for each low-resolution patch. The high-resolution patches are combined to generate a synthesized high-resolution image corresponding to the low-resolution input image.
-
公开(公告)号：US09519975B2
公开(公告)日：2016-12-13
申请号：US14149839
申请日：2014-01-08
Inventor： Yan Wang
CPC classification number: G06T7/0085 , G06F3/038 , G06F3/042 , G06F3/0425 , G06T7/13 , G06T7/174
Abstract: A method of determining an edge of an object on a digital image sequence comprising the step of determining a first gradient direction profile of a first image in the digital image sequence; determining a second gradient direction profile of a second image in the digital image sequence; computing a differential profile based on the first gradient direction profile and the second gradient direction profile; and determining the edge of the object based on the differential profile wherein the differential profile registers gradient magnitudes of the second gradient direction profile and angular differences between the first gradient direction profile and the second gradient direction profile. A system thereof is also disclosed.
Abstract translation: 一种确定数字图像序列上的对象边缘的方法，包括确定数字图像序列中的第一图像的第一梯度方向轮廓的步骤; 确定所述数字图像序列中的第二图像的第二梯度方向轮廓; 基于所述第一梯度方向轮廓和所述第二梯度方向轮廓来计算差分轮廓; 以及基于所述差分轮廓来确定所述对象的边缘，其中所述差分轮廓记录所述第二梯度方向轮廓的梯度大小和所述第一梯度方向轮廓与所述第二梯度方向轮廓之间的角度差。 还公开了其系统。
-
4.发明申请公开(公告)号：US20200151852A1
公开(公告)日：2020-05-14
申请号：US16185860
申请日：2018-11-09
Inventor： Yan Wang , Hailiang Li , Yang Liu , Man Yau Chiu , Zhi Bin Lei
Abstract: Methods and systems which provide super-resolution synthesis based on weighted results from a random forest classifier are described, Embodiments apply a trained random forest classifier to low resolution patches generated from the low-resolution input image to classify the low resolution input patches. As each low-resolution patch is fed into the random forest classifier, each decision tree in the random forest classifier “votes” for a particular class for each of the low-resolution patches. Each class is associated with a projection matrix. The projection matrices output by the decision trees are combined by a weighted average to calculate an overall projection matrix corresponding to the random forest classifier output, which is used to calculate a high-resolution patch for each low-resolution patch. The high-resolution patches are combined to generate a synthesized high-resolution image corresponding to the low-resolution input image.
-
-
-
Search Results
4
records
(took 0.021 seconds)
