Measuring the absolute distance of a front vehicle from an in-car camera based on monocular vision and instance segmentation
Measuring the absolute distance of a front vehicle from an in-car camera based on monocular vision and instance segmentation
复制标题
基于单目视觉和实例分割测量前方车辆与车载摄像头的绝对距离
DOI：
10.1117/1.jei.27.4.043019
复制
发表时间：
2018-07
影响因子：
1.1
通讯作者：
Chen Z
中科院分区：
文献类型：
--
作者：
Huang L;Chen Y;Fan Z;Chen Z
抽象。随着汽车驾驶技术的发展，对汽车测距的要求也越来越高。传统的距离估计方法需要对摄像机的内外参数进行复杂的标定。目前基于神经网络结构的方法主要是测量整个图像的相对深度。我们采用单目视觉与实例分割和摄像机焦距检测的绝对距离的前汽车车内摄像机。首先，我们从目标检测网络中提取汽车的位置。其次，将汽车的位置发送到车辆分类网络和实例分割网络，以获得汽车的类型及其掩码值。在这里，我们使用CompCars数据集训练的模型来分类汽车类型，并使用Cityscapes数据集训练一个新的实例分割模型来获得每个汽车的掩码。第三，根据摄像机成像原理，根据不同车型的尺寸信息与其掩模值之间的关系，计算图像中汽车的绝对距离。利用KITTI数据集对该方法进行了验证，实验表明，该方法的结果可以接近地面真实值。此外，该方法使用实例分割网络来降低深度估计过程的复杂度，即使在汽车部分被遮挡的情况下，仍然可以产生令人满意的结果。
Abstract. The requirements of distance measurement have increased with the development of auto vehicle driving. Traditional methods for distance estimation require the complex calibration from intrinsic and external parameters of the camera. Recent methods based on the neural network structure mainly measure the relative depth of whole images. We adopt monocular vision with instance segmentation and camera focal length to detect the absolute distance of front cars from in-car cameras. First, we extract the location of the cars from the object detection network. Second, the location of cars is sent to the vehicle classification network and instance segmentation network to obtain the type of the cars and their mask value. Here, we use a model trained by the CompCars dataset to classify car types, and we train a new instance segmentation model using the Cityscapes dataset to obtain each car’s mask. Third, in accordance with the camera imaging principle, the absolute distance of cars in the images is calculated based on the relationship between the size information of different car types and their mask values. The proposed method is examined with the KITTI dataset, and the experiment shows that its results can be close to the ground truth. Moreover, the proposed method uses the instance segmentation network to reduce complexity of the depth estimation process and it can still generate satisfactory results even when cars are partly occluded.
登录
查看更多内容
影响因子：
1.3
作者：
Faisal Mahmood;N. Durr
通讯作者：
Faisal Mahmood;N. Durr
DOI：
10.1109/tpami.2018.2839602
发表时间：
2018-03
影响因子：
23.6
作者：
Dan Xu;E. Ricci;Wanli Ouyang;Xiaogang Wang;N. Sebe
通讯作者：
Dan Xu;E. Ricci;Wanli Ouyang;Xiaogang Wang;N. Sebe
DOI：
--
发表时间：
2017-04
期刊：
ArXiv
影响因子：
--
作者：
Bo Li;Yuchao Dai;Huahui Chen;Mingyi He
通讯作者：
Bo Li;Yuchao Dai;Huahui Chen;Mingyi He
DOI：
--
发表时间：
2016-04
期刊：
ArXiv
影响因子：
--
作者：
Weifeng Chen;Z. Fu;Dawei Yang;Jia Deng
通讯作者：
Weifeng Chen;Z. Fu;Dawei Yang;Jia Deng
DOI：
10.1109/tpami.2016.2572683
发表时间：
2017-04-01
影响因子：
23.6
作者：
Shelhamer, Evan;Long, Jonathan;Darrell, Trevor
通讯作者：
Darrell, Trevor