公开(公告)号：US20140267348A1

公开(公告)日：2014-09-18

申请号：US14216637

申请日：2014-03-17

Applicant: GOOGLE INC.

Inventor： Alex Gontmakher ,  Andrin von Rechenberg

IPC: G06T5/00

CPC classification number: G06T5/002 ,  G06T17/00 ,  G06T2207/20182 ,  G06T2210/36

Abstract: Methods and systems for fractional level of detail assignment are described herein. A method embodiment for fractional level of detail (LOD) assignment includes obtaining a set of features and image data at a range of LOD values, assigning one or more fractional LOD values to the obtained features and providing the features and the image data at the fractional LOD values. The embodiment also includes hashing an identifier associated with each feature and computing a hash cutoff value by mapping the range of LOD levels onto a range of integers. A system embodiment includes a LOD assigner to assign fractional LOD values to features in image data and to provide the features and the image data at the fractional LOD values. The system embodiment further includes a retrieval engine to return features with a range of LOD values that include the fractional LOD values to the LOD assigner.

Abstract translation: 本文描述了分数级细节分配的方法和系统。 用于分数级别细节（LOD）分配的方法实施例包括在LOD值的范围内获得一组特征和图像数据，将一个或多个分数LOD值分配给所获得的特征，并将分数的特征和图像数据提供 LOD值。 该实施例还包括散列与每个特征相关联的标识符，并且通过将LOD级别的范围映射到整数范围来计算散列截止值。 系统实施例包括LOD分配器，以将分数LOD值分配给图像数据中的特征，并且以分数LOD值提供特征和图像数据。 该系统实施例还包括检索引擎，以向LOD分配器返回具有包括分数LOD值的一系列LOD值的特征。

公开(公告)号：US10121230B2

公开(公告)日：2018-11-06

申请号：US14216637

申请日：2014-03-17

Applicant: GOOGLE INC.

Inventor： Alex Gontmakher ,  Andrin von Rechenberg

IPC: G06T5/00 ,  G06T17/00

Abstract: Methods and systems for fractional level of detail assignment are described herein. A method embodiment for fractional level of detail (LOD) assignment includes obtaining a set of features and image data at a range of LOD values, assigning one or more fractional LOD values to the obtained features and providing the features and the image data at the fractional LOD values. The embodiment also includes hashing an identifier associated with each feature and computing a hash cutoff value by mapping the range of LOD levels onto a range of integers. A system embodiment includes a LOD assigner to assign fractional LOD values to features in image data and to provide the features and the image data at the fractional LOD values. The system embodiment further includes a retrieval engine to return features with a range of LOD values that include the fractional LOD values to the LOD assigner.