Detection of approaching cars via artificial insect vision
Detection of approaching cars via artificial insect vision
复制标题
通过人工昆虫视觉检测接近的汽车
DOI：
10.1002/ecjc.20195
复制
发表时间：
2005
期刊：
影响因子：
--
通讯作者：
Kazutaka Niijima
中科院分区：
文献类型：
--
作者：
N. Ohta;Kazutaka Niijima
为了帮助汽车驾驶员，使用图像处理来检测从左后方或右后方接近的另一辆车，这可能很容易进入盲点。用于检测的基本信息是图像运动（光流）。然而，通过传统手段计算图像运动是操作密集型的，并且需要昂贵的设备。然而，在注意到昆虫使用来自简单传感器（如复眼）的信息和由简单神经系统执行的处理来获得足够的信息之后，作者试图根据简单处理和低分辨率图像信息来达到他们的目的。虽然他们使用普通的相机和计算机进行实验，但作者认为该技术未来将在通过使用简单的电子电路和光电二极管阵列等光接收元件创建的设备上实现。所提出的技术的一个主要特点是，它的计算复杂度低。与传统技术的平均图像处理相比，计算时间的差异约为10 - 5。© 2005威利期刊有限公司Electron Comm Jpn Pt 3，88（10）：57-65，2005;在线发表于Wiley InterScience（www.interscience.wiley.com）。DOI 10.1002/ecjc.20195
To aid an automobile driver, image processing is used to detect another vehicle approaching from the left or right rear, which may easily enter a blind spot. The basic information that is used for detection is image movement (optical flow). However, calculating image movement by traditional means is operationally intensive and requires expensive equipment. However, after noticing that an insect uses information from simple sensors such as compound eyes and processing performed by a simple nervous system to obtain sufficient information, the authors tried to reach their objective according to simple processing and low-resolution image information. Although they used a normal camera and computer for the experiments, the authors assume that the technique will be implemented in the future on a device created by using simple electronic circuits and light-receiving elements such as photodiode arrays. A major feature of the proposed technique is that its computational complexity is low. The difference in computation time compared with the average image processing for the conventional technique was on the order of 10−5. © 2005 Wiley Periodicals, Inc. Electron Comm Jpn Pt 3, 88(10): 57–65, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecjc.20195