Recognition on Images From Internet Street View Based on Hierarchical Features Learning With CNNs

Recognition on Images From Internet Street View Based on Hierarchical Features Learning With CNNs

Jian-min Liu (Central South University, China & Hunan Institute of Humanities, Science and Technology, China) and Min-hua Yang (Central South University, China)
DOI: 10.4018/978-1-5225-8054-6.ch062
OnDemand PDF Download:
No Current Special Offers


This article describes hierarchical features with unsupervised learning on images from internet street view images. This is due to the time spent by trained researchers on feature construction steps with traditional methods. This article focuses on the activation of each layer of with convolutional neural networks (CNNs) on Internet street view images detection and compared similarities and differences among them on each layer. The experiment results achieved error rates of 21% on recognition which work went relatively well than the traditional machine learning techniques, such as Parallel SVM.
Chapter Preview

Traditional image processing approach work quite well on low resolution remote sensing images(LRRS) but obtain low hit rate high resolution remote sensing images (HRRS). Depend on features template (Zhanga & Zhou, 2004), the researcher achieve image and template matching, and high hit rate of object recognition.

Object recognition of HRRS can depend on intra-domain knowledge provided by experienced experts (Durand, Derivaux, Forestier, Wemmert, Gançarski, Boussaid & Puissant, 2007). The results of experiments show relatively good results and high hit rate when distinguish objects in urban and rural areas.

Different kernel-based approaches of hyperspectral image classification are compared with each other, which include kernel fisher discriminant(KFD), regularized radial basis function neural networks(Reg-RBFNN), regularized AdaBoost (Reg-AB), standard support vector machines (SVMs). The results show that SVMs achieve better results than other several kernel-based approaches and at a much less time in most cases. (Gustavo Camps-Valls, Lorenzo Bruzzone, 2005).

Complete Chapter List

Search this Book: