A realistic population density distribution scenario in conjunction with the spatial dynamic spectrum allocation （DSA） is taken into account to mitigate the spectrum wastage in terms of extra guard bands. For the in...A realistic population density distribution scenario in conjunction with the spatial dynamic spectrum allocation （DSA） is taken into account to mitigate the spectrum wastage in terms of extra guard bands. For the insertion of the extra guard bands, an efficient strategy based on self-assessment is applied to each victim cell individually and independently. Consequently, it is no more required to spread the extra guard band over the whole DSA region. Simulation results StlOW an improvement of 3% -4% in percentage of satisfied users for Universal Mobile Telecommunications System （UMTS） network and 4%-5% for Digital Video Broadcasting Terrestrial （DVB-T） network.展开更多
车道线检测是保证自动驾驶安全性与稳定性的关键,为提高车道线检测的准确性,本文基于UFLD(Ultra Fast Structure-aware Deep Lane Detection)算法,结合DenseNet-121网络和空间注意力(Spatial Attention)机制,设计了一种DSA-UFLD模型实...车道线检测是保证自动驾驶安全性与稳定性的关键,为提高车道线检测的准确性,本文基于UFLD(Ultra Fast Structure-aware Deep Lane Detection)算法,结合DenseNet-121网络和空间注意力(Spatial Attention)机制,设计了一种DSA-UFLD模型实现车道线检测。在图像增强方面,使用图像亮度自适应增强算法提高欠曝图像的清晰度;在网络优化方面,用迁移学习模型DenseNet-121代替ResNet18提取图像特征,利用密集连接加强特征重用,并引入空间注意力机制提取图像的关键信息,其次在上采样中用转置卷积代替双线性插值,通过学习参数,更好地实现解码;在损失函数方面,通过改进结构损失,将车道线约束为二次曲线,改善了弯道场景下车道线的检测效果。实验结果表明,DSA-UFLD算法在保证检测速度的同时,提高了车道线的识别准确率,具有一定的应用价值。展开更多
基金The National High-Tech Research and Development Program of China ( No.2005AA123950)the National Science Foundation of China (No.90604035)
文摘A realistic population density distribution scenario in conjunction with the spatial dynamic spectrum allocation （DSA） is taken into account to mitigate the spectrum wastage in terms of extra guard bands. For the insertion of the extra guard bands, an efficient strategy based on self-assessment is applied to each victim cell individually and independently. Consequently, it is no more required to spread the extra guard band over the whole DSA region. Simulation results StlOW an improvement of 3% -4% in percentage of satisfied users for Universal Mobile Telecommunications System （UMTS） network and 4%-5% for Digital Video Broadcasting Terrestrial （DVB-T） network.
文摘车道线检测是保证自动驾驶安全性与稳定性的关键,为提高车道线检测的准确性,本文基于UFLD(Ultra Fast Structure-aware Deep Lane Detection)算法,结合DenseNet-121网络和空间注意力(Spatial Attention)机制,设计了一种DSA-UFLD模型实现车道线检测。在图像增强方面,使用图像亮度自适应增强算法提高欠曝图像的清晰度;在网络优化方面,用迁移学习模型DenseNet-121代替ResNet18提取图像特征,利用密集连接加强特征重用,并引入空间注意力机制提取图像的关键信息,其次在上采样中用转置卷积代替双线性插值,通过学习参数,更好地实现解码;在损失函数方面,通过改进结构损失,将车道线约束为二次曲线,改善了弯道场景下车道线的检测效果。实验结果表明,DSA-UFLD算法在保证检测速度的同时,提高了车道线的识别准确率,具有一定的应用价值。
