期刊文献+

面向移动平台的深度学习复杂场景目标识别应用 被引量:4

A deep learning based object detection application for mobile platform in complex scenes
下载PDF
导出
摘要 针对传统建筑物提取方法对人为设计的依赖,以及对建筑物边缘特征提取算法的改进,通过Keras框架获取卷积神经网络(convolutional neural networks,CNN)模型MobileNet的瓶颈层后加入新的分类器进行迁移学习,对输入图片进行大量的图像增强技术和测试集增强技术,经过三个阶段的迁移学习后获得了较高的准确率。相比其他的特征提取算法,CNN具有平移不变性以及自动提取特征等优点,在较短的时间内获得较高准确率的同时,MobileNet的权重仅有15.3 MB,兼顾计算量和精度,可以广泛移植到移动端设备。基于模型移植的移动端系统兼具拍照识别、相册识别、菜单展示等功能,为移动平台用户快速准确地判断自然场景中建筑物的信息提供了便捷工具。 Due to the presence of background noise in natural scenes and the interference of complex factors such as illumination, rotation, and shooting angle, it is very difficult to identify the image of buildings in natural scenes. Aiming at the dependence of traditional building extraction methods on human design and the improvement of building edge feature extraction algorithm.Through the Keras framework to obtain the bottleneck layer of convolutional neural networks(CNN) model MobileNet,and add a new classifier for transfer learning. A large number of data augmentation and test set augmentation are applied to the input image. After three versions of transfer learning, high accuracy was achieved within 480 iterations in three test set. Compared with other feature extraction algorithms, CNN has the advantages of non-transformation and automatic extraction of features, achieves higher accuracy in a shorter period of time. At the same time, MobileNet weight only occupy 15.3 MB with high precision and less calculation, which can be widely transplanted to mobile devices. The system based on model migration has the functions of photo recognition, photo album recognition, menu display, etc., providing mobile platform users with a convenient and simple tool to quickly and accurately obtain the information of buildings in natural scenes.
作者 许博鸣 刘晓峰 业巧林 张福全 周京正 XU Boming;LIU Xiaofeng;YE Qiaolin;ZHANG Fuquan;ZHOU Jingzheng(College of Information Science and Technology, Nanjing Forestry University, Nanjing 210037, Jiangsu, China;Bureau of Information Technology, Ministry of Public Security of the People′s Republic of China, Beijing 100741, China)
出处 《陕西师范大学学报(自然科学版)》 CAS CSCD 北大核心 2019年第5期10-15,共6页 Journal of Shaanxi Normal University:Natural Science Edition
基金 国家自然科学基金(61871444,31670554) 南京林业大学大学生创新训练计划项目(2017NFUSPITP231)
关键词 迁移学习 深度学习 卷积神经网络 移动平台移植 人工智能 transfer learning deep learning convolutional neural network mobile system transplantation artificial intelligence
  • 相关文献

参考文献5

二级参考文献29

  • 1王守觉,曹文明.半导体神经计算机的硬件实现及其在连续语音识别中的应用[J].电子学报,2006,34(2):267-271. 被引量:3
  • 2D M McKeown.Toward automatic cartographic feature extractionIn:Mapping and Spatial Modelling for Navigation[C].NATO ASI Series.Berlin:Springer-Verlag,1990,F65:149-180
  • 3R B Irvinm,D M McKeown.Methods for exploiting the relationship between buildings and their shadows in aerial imagery[J].IEEE Trans on Systems,Man,and Cybernetics,1989,19(6):1564-1575
  • 4J C McGlone,J A Shufelt.Projective and object space geometry for monocular building extraction[C].IEEE Conf on Computer Vision and Pattern Recognition,Washington,USA,1994
  • 5J A Shufelt.Exploiting photogrammetric methods for building extraction in aerial images[J].Int'l Archives of Photogrammetry and Remote Sensing,1996,31(B6/S):74-79
  • 6J A Shufelt.Projective geometry and photometry for object detection and delineation[R].Carnegie Mellon University,Tech Rep:CMU-CS-96-164,1996
  • 7C Lin,R Nevatia.Building detection and description from a single intensity image[J].Computer Vision and Image Understanding,1998,72(2):101-121
  • 8Yunqi Song,Aidong Zhang.Analyzing scenery images by monotonic tree[J].ACM Multimedia Systems Journal,2002,8(6):495-511
  • 9A Iqbal,J K Aggarwal.Applying perceptual grouping to content-based image retrieval:Building images[C].IEEE Int'l Conf on CVPR,Fort lolins,lolorado,1999
  • 10S Kumar,M Hebert.Man-made structure detection in natural images using a causal multiscale random field[C].IEEE Int'l Conf on CVPR,Madison,USA,2003

共引文献2194

同被引文献36

引证文献4

二级引证文献18

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部