公开(公告)号：US09399463B2

公开(公告)日：2016-07-26

申请号：US14582596

申请日：2014-12-24

Applicant: Industrial Technology Research Institute

Inventor： Long-Der Chen ,  Dau-Chen Huang ,  Pai-Wei Cheng ,  Ke-Horng Chen ,  Shin-Chi Lai ,  Kuo-Chen Kao ,  Sheau-Fang Lei

IPC: G06F7/70 ,  B60W30/08 ,  B60T7/22

CPC classification number: B60W30/08 ,  B60T7/22 ,  B60T2201/022 ,  B60T2210/32 ,  B60W10/18 ,  B60W30/09 ,  B60W50/14 ,  B60W2050/143 ,  B60W2050/146 ,  B60W2420/42 ,  B60W2550/10 ,  B60W2550/302 ,  B60W2550/308

Abstract: An automatic tracking collision avoidance method has following steps: taking images in front of an origin vehicle by at least two photograph units; the images delivering to a calculation assembly, to transform the images from three primary colors images to grey images; having a relative vehicle speed of the front vehicle corresponded to the original vehicle according to the grey images; and according to the grey images the calculation assembly having a relative vehicle speed that the front vehicle corresponded to the original vehicle; the calculation assembly having a safe distance according to the relative vehicle speed, if the safe distance is great than a distance that between the original vehicle and the front vehicle, an alarm module generating an alert signal.

Abstract translation: 自动跟踪碰撞避免方法具有以下步骤：通过至少两张照片单元拍摄原始车辆前方的图像; 将图像传送到计算组件，以将图像从三原色图像转换成灰度图像; 根据灰度图像具有前车辆的相对车速对应于原车辆; 并且根据所述灰色图像，所述计算组件具有所述前部车辆对应于所述原始车辆的相对车辆速度; 如果安全距离大于原车和前车之间的距离，则计算组件根据相对车速具有安全距离，产生报警信号的报警模块。

公开(公告)号：US09918010B2

公开(公告)日：2018-03-13

申请号：US14977793

申请日：2015-12-22

Applicant: INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE

Inventor： Pai-Wei Cheng ,  Long-Tai Chen ,  Po-Feng Chang

IPC: H04N7/00 ,  H04N5/232 ,  H04N17/00 ,  G06K9/00 ,  G06T5/50 ,  B60R1/00 ,  H04N5/247 ,  G06T7/33

CPC classification number: H04N5/23238 ,  B60R1/00 ,  B60R2300/105 ,  B60R2300/402 ,  G06K9/00791 ,  G06K9/46 ,  G06T3/4038 ,  G06T5/50 ,  G06T7/33 ,  G06T7/80 ,  G06T2207/30252 ,  H04N5/247 ,  H04N17/002

Abstract: A method for adjusting vehicle panorama system, which comprises: a calibration pattern is disposed around a vehicle, and at least one patterned feature and a plurality of cameras are disposed on the vehicle; capture the calibration pattern and patterned feature to obtain an original image; capture the patterned feature after a period of time to obtain a correction image; compares the correction image and original image and corrects the panorama images by recombining the images shoot by the plurality of cameras according to a comparison result.