公开(公告)号：US20180137647A1

公开(公告)日：2018-05-17

申请号：US15813668

申请日：2017-11-15

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Jia LI ,  Qiang WANG ,  Feng SHI ,  Deheng QIAN ,  Dongqing ZOU ,  Hyunsurk Eric RYU ,  JinMan PARK ,  Jingtao XU ,  Keun Joo PARK ,  Weiheng LIU

IPC: G06T7/73 ,  G06K9/66 ,  G06T7/11

CPC classification number: G06T7/73 ,  G06K9/00805 ,  G06K9/4628 ,  G06K9/4671 ,  G06K9/6274 ,  G06K9/66 ,  G06N3/0445 ,  G06T7/11 ,  G06T2210/12

Abstract: The disclosure provides an object detection method and apparatus based on a Dynamic Vision Sensor (DVS). The method includes the following operations of: acquiring a plurality of image frames by a DVS; and, detecting the image frames by a recurrent coherent network to acquire a candidate box for objects to be detected, wherein the recurrent coherent network comprising a frame detection network model and a candidate graph model. By using a new recurrent coherent detection network, a bounding box for an object to be detected is fast detected from the data acquired by a DVS. The detection speed is improved greatly while ensuring the detection accuracy.

公开(公告)号：US20180137639A1

公开(公告)日：2018-05-17

申请号：US15812554

申请日：2017-11-14

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Dongqing ZOU ,  Feng SHI ,  Weiheng LIU ,  Deheng QIAN ,  Hyunsurk Eric RYU ,  Jia LI ,  Jingtao XU ,  Keun Joo PARK ,  Qiang WANG ,  Changwoo SHIN

IPC: G06T7/593 ,  H04N13/00

CPC classification number: G06T7/593 ,  G06T2207/10012 ,  G06T2207/10028 ,  H04N13/128 ,  H04N13/239 ,  H04N2013/0081

Abstract: Exemplary embodiments provide an image vision processing method, device and equipment and relate to: determining parallax and depth information of event pixel points in a dual-camera frame image acquired by Dynamic Vision Sensors; determining multiple neighboring event pixel points of each non-event pixel point in the dual-camera frame image; determining, according to location information of each neighboring event pixel point of each non-event pixel point, depth information of the non-event pixel point; and performing processing according to the depth information of each pixel point in the dual-camera frame image. Since non-event pixel points are not required to participate in the matching of pixel points, even if it is difficult to distinguish between the non-event pixel points or the non-event pixel points are occluded, depth information of the non-event pixel points can be accurately determined according to the location information of neighboring event pixel points.

公开(公告)号：US20160320834A1

公开(公告)日：2016-11-03

申请号：US15017711

申请日：2016-02-08

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Jun Haeng LEE ,  Keun Joo PARK ,  Changwoo SHIN ,  Jooyeon WOO ,  Hyunsurk Eric RYU ,  Seokjin HONG

IPC: G06F3/01 ,  G06T7/20 ,  G06F1/16 ,  G06F3/16 ,  G06F1/32 ,  G06K9/46 ,  G06T7/00

CPC classification number: G06F3/012 ,  G06F1/1633 ,  G06F1/3203 ,  G06F3/017 ,  G06F3/167 ,  G06K9/00355 ,  G06K9/00389 ,  G06K9/4604 ,  G06T7/20 ,  G06T7/246 ,  G06T2207/10028

Abstract: A method of extracting a static pattern from an output of an event-based sensor. The method may include receiving an event signal from the event-based sensor in response to dynamic input, and extracting a static pattern associated with the dynamic input based on an identifier and time included in the event signal. The static pattern may be extracted from a map generated based on the identifier and time.

Abstract translation: 从基于事件的传感器的输出中提取静态图案的方法。 该方法可以包括响应于动态输入从基于事件的传感器接收事件信号，以及基于包括在事件信号中的标识符和时间来提取与动态输入相关联的静态模式。 可以从基于标识符和时间生成的映射中提取静态模式。