-
公开(公告)号：US10863142B2
公开(公告)日：2020-12-08
申请号：US16139527
申请日：2018-09-24
Inventor： Tomoaki Itoh , Hidehiko Shin
IPC: H04N7/18 , G08B13/196 , G06T7/254 , G06K9/00 , G06T1/00
Abstract: A monitoring device (2) identifies an object from videos made by a plurality of cameras (1) including a first camera and a second camera and having a predetermined positional relationship. The monitoring device has a receiving unit (21) configured to receive the videos from the plurality of cameras, a storage unit (22b, 22c) configured to store feature information indicating a feature of the object and camera placement information indicating placement positions of the cameras, and a controller (23) configured to identify the object from the videos based on the feature information. If an object has been identifiable from the video made by the first camera but has been unidentifiable from the video made by the second camera, the controller (23) specifies, based on the camera placement information, the object in the video made by the second camera.
-
公开(公告)号：US12125274B2
公开(公告)日：2024-10-22
申请号：US17485841
申请日：2021-09-27
Inventor： Tomoaki Itoh , Hidehiko Shin
CPC classification number: G06V10/987 , G06N20/00 , G06V10/25 , G06V10/94
Abstract: It reduces labor and time to generate training data for the training model. An identification information assignment apparatus includes an acquirer configured to acquire a plurality of pieces of image data, an assigner configured to assign identification information to image data selected from the plurality of pieces of image data by using a learning model after learning, and an updater configured to update the learned model using the image data to which the identification information is assigned, wherein the assigner assigns identification information to a rest of the image data acquired by the acquirer using the learned model that has been updated.
-
公开(公告)号：US20240112499A1
公开(公告)日：2024-04-04
申请号：US18536634
申请日：2023-12-12
Inventor： Tomoaki Itoh , Tomokazu Ichiriki , Hidehiko Shin , Daijiroh Ichimura
Abstract: An image analysis device comprising: an input interface to acquire image data indicating an image of an operator performing task; and a controller to detect the task performed by the operator sequentially, based on the image data, to generate task detection information indicating a detection result of the task, wherein the task includes a plurality of processes performed by the operator, and the controller compensates the task detection information by interpolating a detection result of the task within an undetected period, in accordance with a pre process and/or a post process of the undetected period, the undetected period being a period in which the task is not detected, the pre process being a process of the task detected previously to the undetected period, and the post process being a process of the task detected subsequently to the undetected period.
-
公开(公告)号：US10271082B2
公开(公告)日：2019-04-23
申请号：US15285736
申请日：2016-10-05
Inventor： Toshiyasu Sugio , Youji Shibahara , Yuki Maruyama , Toru Matsunobu , Yoichi Sugino , Mikihiro Ouchi , Hisao Sasai , Kuniaki Isogai , Ryuji Muta , Takako Hori , Tomoaki Itoh
IPC: H04N21/262 , H04N21/218 , H04N21/2187 , H04N21/2665 , H04N7/18 , H04N21/234
Abstract: A video distribution method is a video distribution method to be performed by a server that distributes a plurality of videos captured by a plurality of users from different viewpoints to a terminal apparatus. The video distribution method includes distributing, from a server to a terminal apparatus, a first video that is one of a plurality of videos captured by a plurality of users from different viewpoints and that is requested by the terminal apparatus, selecting a second video that is one of the plurality of videos and that is likely to be next requested next time by the terminal apparatus, and starting transmission of the second video to the terminal apparatus during distribution of the first video to the terminal apparatus.
-
-
-
Search Results
4
records
(took 0.017 seconds)
