Automotive Research & Testing Center

Chuan-Ren Li

Kuo-Ching Chang

Han-Wen Huang

Li-You Syu

Zhen-Wei Zhu

Yun-Ling Chang

A method is to be implemented by a processor connected to a storage device and an image capturing device, and includes steps of: performing a first and a second classifications on an image received from the image capturing device for respectively obtaining first and second results of obstacle detection, and storing the same in the storage device; determining whether there is a distinct obstacle, which is an obstacle indicated by one of the first and second results of obstacle detection, and not indicated by the other one of the first and second results of obstacle detection; and calculating a penalty score based on the distinct obstacle, and updating a credibility score by subtracting the penalty score therefrom.

Method for evaluating credibility of obstacle detection

A hierarchical system for detecting an object with parallel architecture and a hierarchical method thereof is disclosed. The system includes at least one image-retrieving device retrieving at least an image and searching a plurality of obstacle position images in it. The image-retrieving device is electrically connected with an image-processing device to receive the obstacle position images transmitted by the image-retrieving device, uses parallel architecture classification to obtain at least one object image and a plurality of cropping frames thereof from the obstacle position images, synchronously separates the cropping frames to retrieve characteristic values of each cropping frame, uses convolutional neural network to simultaneously recognize the characteristic values of each cropping frame, and searches and outputs the correct cropping frame from the object image, thereby immediately detecting the object outside a vehicle and obtaining the cropping frame of the object to avoid detection error.

Hierarchical system for detecting object with parallel architecture and hierarchical method thereof

A method for image-based dynamic feature enhancement includes: using the camera to capture multiple images, selecting an ROI from each of the multiple images displayed on the camera through the ROI selection unit, configuring multiple brightness weights to correspond to the ROI, adjusting brightness of the image according to the multiple brightness weights and multiple actual brightness values corresponding to the ROI, and performing a multi-layer progressive sharpening process to adjust multiple actual sharpening parameters of the image through the sharpening adjustment unit. The camera parameters of the camera can be directly and instantly adjusted to improve brightness and sharpening of images captured by the camera and facilitate advanced driver assistance system to clearly identify road conditions.

Method and system for dynamically adjusting parameters of camera settings for image enhancement

A target object detection system includes a detection unit and a machine learning unit. The detection unit obtains an image and outputs a notification signal upon determining that a portion of the image resembles a target object by performing a classification operation. The machine learning unit performs a training operation using the image to obtain new weak classifier parameters for update in the detection unit when a determination made using weak classifier parameters used in the classification operation of the detection unit differs from one made using strong classifier parameters.

Growing community of inventors

Lugong, Taiwan

Chuan-Ren Li