A neural network approach using multi-scale textural metrics from very high-resolution panchromatic imagery for urban land-use classification

<p><strong>Abstract.</strong> Rapid urbanization of Indian cities requires a focused attention with respect to preparation of Master Plans of cities. Urban land use/land cover from very high resolution satellite data sets is an important input for the preparation of the master plans of the cities along with extraction of transportation network, infrastructure details etc. Conventional classifiers, which are pixel based do not yield reasonably accurate urban land use/land cover classification of very high resolution satellite data (usually merged images of Panchromatic &amp; Multispectral). Object Based Image Classification techniques are being used to generate urban land use maps with ease which is GIS compatible while using very high resolution satellite data sets. In this study, Object Based Image Analysis (OBIA) has been used to create broad level urban Land Use / Land Cover (LU/LC) map using high resolution ResourceSat-2 LISS-4 and Cartosat-1 pan-sharpened image on the study area covering parts of East Delhi City. Spectral indices, geometric parameters and statistical textural methods were used to create algorithms and rule sets for feature classification. A LU/LC map of the study area comprising of 4 major LU/LC classes with its main focus on separation of barren areas from built up areas has been attempted. The overall accuracy of the result obtained is estimated to be approximately 70%.</p>

Download Full-text

Integrating Aerial and Street View Images for Urban Land Use Classification

Remote Sensing ◽

10.3390/rs10101553 ◽

2018 ◽

Vol 10 (10) ◽

pp. 1553 ◽

Cited By ~ 28

Author(s):

Rui Cao ◽

Jiasong Zhu ◽

Wei Tu ◽

Qingquan Li ◽

Jinzhou Cao ◽

...

Keyword(s):

Neural Network ◽

Land Use ◽

Classification Accuracy ◽

Deep Neural Network ◽

Ground Level ◽

Urban Land ◽

Urban Land Use ◽

Aerial Images ◽

Land Use Classification ◽

Street View

Urban land use is key to rational urban planning and management. Traditional land use classification methods rely heavily on domain experts, which is both expensive and inefficient. In this paper, deep neural network-based approaches are presented to label urban land use at pixel level using high-resolution aerial images and ground-level street view images. We use a deep neural network to extract semantic features from sparsely distributed street view images and interpolate them in the spatial domain to match the spatial resolution of the aerial images, which are then fused together through a deep neural network for classifying land use categories. Our methods are tested on a large publicly available aerial and street view images dataset of New York City, and the results show that using aerial images alone can achieve relatively high classification accuracy, the ground-level street view images contain useful information for urban land use classification, and fusing street image features with aerial images can improve classification accuracy. Moreover, we present experimental studies to show that street view images add more values when the resolutions of the aerial images are lower, and we also present case studies to illustrate how street view images provide useful auxiliary information to aerial images to boost performances.

Download Full-text

A detailed comparison of backpropagation neural network and maximum-likelihood classifiers for urban land use classification

IEEE Transactions on Geoscience and Remote Sensing ◽

10.1109/36.406684 ◽

1995 ◽

Vol 33 (4) ◽

pp. 981-996 ◽

Cited By ~ 237

Author(s):

J.D. Paola ◽

R.A. Schowengerdt

Keyword(s):

Neural Network ◽

Land Use ◽

Maximum Likelihood ◽

Detailed Comparison ◽

Urban Land ◽

Urban Land Use ◽

Land Use Classification ◽

Backpropagation Neural Network

Download Full-text