-
公开(公告)号：US11037016B2
公开(公告)日：2021-06-15
申请号：US16391621
申请日：2019-04-23
发明人： Yen-Lin Chen , Hong-Yi Liang , Xiu-Zhi Chen , Chao-Wei Yu , Meng-Tsan Li
摘要： A feature determination apparatus and method adapted to multiple object sizes are provided. The apparatus individually supplies each of the object images to a convolution neural network having several convolution layers to generate multiple feature maps corresponding to each object image. The apparatus calculates a feature amount of each feature image of each object image. The apparatus determines an invalid layer start number of each object image according to a preset threshold and the feature amount corresponding to each object image. The apparatus determines a feature map extraction recommendation for each of a plurality of object sizes according to a size of each object image and the invalid layer start number of each object image.
-
2.发明授权公开(公告)号：US10430663B2
公开(公告)日：2019-10-01
申请号：US15834510
申请日：2017-12-07
发明人： Hsin-Han Chiang , Yen-Lin Chen , Chien Lin , Chao-Wei Yu , Meng-Tsan Li
摘要： An image annotation method includes the following. Image frames are acquired. One or more target objects are identified and tracked from the image frames. Candidate key image frames are selected from the frames according to a first selection condition. First similarity indexes of the candidate key image frames are determined. Second similarity indexes of a plurality of adjacent frames are determined. The candidate key image frames as well as the adjacent frames that meet a second selection condition are selected as key image frames. The key image frames are displayed on a graphical user interface. Annotation information about the one or more target objects is display through the graphical user interface.
-
公开(公告)号：US09898672B2
公开(公告)日：2018-02-20
申请号：US15084465
申请日：2016-03-29
发明人： Cheng-Lung Jen , Yen-Lin Chen , Chao-Wei Yu , Meng-Tsan Li , Augustine Tsai
CPC分类号： G06K9/00825 , B60Q9/00 , G06K9/4642 , G06K9/4652 , G06K9/4661 , G06K9/468 , G06K2209/23 , G06T7/246 , G06T7/251 , G06T7/277 , G06T2207/20016 , G06T2207/30252
摘要： A system of detection, tracking and identification of an evolutionary adaptation of a vehicle lamp includes an image capture device and a processor. The image capture device captures an image of a vehicle. The processor processes the image of the vehicle to generate a detection result of the vehicle lamp, analyzes and integrates vehicle lamp dynamic motion information and vehicle lamp multiple scale variation information based on the detection result, and then tracks the position of the vehicle lamp by applying a multiple scale vehicle lamp measurement model.
-
公开(公告)号：US10684357B2
公开(公告)日：2020-06-16
申请号：US15787535
申请日：2017-10-18
发明人： Hsin-Han Chiang , Yen-Lin Chen , Chien Lin , Chao-Wei Yu , Meng-Tsan Li
摘要： A vehicle detecting detection assisting method and a vehicle detection assisting system are provided. The vehicle detection assisting method includes the following steps. A scanning range of a lidar unit is obtained. A width of a lane is obtained. A trace of the lane is obtained. A dynamic region of interest in the scanning range is created according to the width and the trace.
-
5.发明授权公开(公告)号：US11270150B2
公开(公告)日：2022-03-08
申请号：US16701140
申请日：2019-12-02
发明人： Yen-Lin Chen , Hsiu-Chih Chen , Chieh-Min Chang , Chao-Wei Yu , Meng-Tsan Li
摘要： A computing device divides a training image into a plurality of training blocks, and the training image includes a training object. The computing device calculates, for each of the training blocks, a correct confidence score of the training object covering the training block according to an image-marking data and a confidence-score-translating function, and the image-marking data includes a piece of location information of the training object in the training image. Then, the computing device trains a deep-learning model with the training image, the correct confidence scores and the image-marking data to generate the object-detecting model.
-
公开(公告)号：US09361544B1
公开(公告)日：2016-06-07
申请号：US14562787
申请日：2014-12-08
发明人： Yen-Lin Chen , Chuan-Yen Chiang , Chao-Wei Yu , Augustine Tsai , Meng-Tsan Li
CPC分类号： G06T3/20 , G06K9/628 , G06K9/6286
摘要： A multi-class object classifying method and system are disclosed herein, where the multi-class object classifying method includes the following steps: classes, first training images and second training images are received and stored, and first characteristic images and second characteristic images are respectively extracted from the first training images and the second training images; the first training images is used to generate classifiers through a linear mapping classifying method; a classifier and the second characteristic images are used to determine parameter ranges corresponding to the classes and a threshold corresponding to the classifier. When two of the parameter ranges overlap, the remaining parameter ranges except for the two overlapped parameter ranges are recorded; after another classifier is selected from the classifiers except for the classifier that has been selected, the previous steps is repeated until the parameter ranges don't overlap with each other and the parameter ranges are recorded.
摘要翻译： 本文公开了一种多类对象分类方法和系统，其中多类对象分类方法包括以下步骤：分类，第一训练图像和第二训练图像被接收和存储，并且第一特征图像和第二特征图像分别 从第一训练图像和第二训练图像中提取; 第一训练图像用于通过线性映射分类方法生成分类器; 分类器和第二特征图像用于确定与类别相对应的参数范围和对应于分类器的阈值。 当两个参数范围重叠时，记录除两个重叠参数范围之外的其余参数范围; 在除了已经选择的分类器之外的分类器中选择另一个分类器之后，重复前面的步骤，直到参数范围彼此不重叠并且记录参数范围。
-
-
-
-
-
搜索结果
6 条记录 (耗时 0.015 秒)
