Innovative Multi Pcnn Based Network for Green Area Monitoring - Identification and Description of Nearly Indistinguishable Areas - In Hyperspectral Satellite Images

The paper presents an original neural network approach for region of interest detection and classification in multi-spectral satellite images. The proposed method uses a sequence of Pulse Coupled Neural Networks that identifies plausible regions of interest. These regions are passed to a dimension reduction algorithm, Principle Component Analysis, in order to generate the input data for a Support Vector Machine classifier, that validates the data. The algorithm's parameters are optimized using a Genetic Algorithm. The algorithm is designed to distinguish regions that are extremely similar, such as parks in a city that has entire districts made up of houses with yards. The algorithm has been tested on images provided by the Sentinel-2 satellite, and it proved that it can recall 76.85% of the pixels marked as park in the ground truth data, which was obtained from Open Street Map.

[1]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  F. J. García-Haro,et al.  A generalized soil-adjusted vegetation index , 2002 .

[3]  Sam Kwong,et al.  Genetic algorithms: concepts and applications [in engineering design] , 1996, IEEE Trans. Ind. Electron..

[4]  Victor-Emil Neagoe,et al.  A new multispectral pixel change detection approach using pulse-coupled neural networks for change vector analysis , 2016, 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[5]  Jinsong Deng,et al.  PCA‐based land‐use change detection and analysis using multitemporal and multisensor satellite data , 2008 .

[6]  Heather McNairn,et al.  International Journal of Applied Earth Observation and Geoinformation , 2014 .

[7]  J. L. Johnson.  Pulse-coupled neural nets: translation, rotation, scale, distortion, and intensity signal invariance for images. , 1994, Applied optics.

[8]  Reinhard Eckhorn,et al.  Feature Linking via Synchronization among Distributed Assemblies: Simulations of Results from Cat Visual Cortex , 1990, Neural Computation.

[9]  Michael Berger,et al.  Sentinel-2 optical high resolution mission for GMES operational services , 2007, 2007 IEEE International Geoscience and Remote Sensing Symposium.

[10]  Shan Suthaharan,et al.  Support Vector Machine , 2016 .

[11]  Peng Gong,et al.  Mapping Urban Land Use by Using Landsat Images and Open Social Data , 2016, Remote. Sens..

[12]  William Stafford Noble,et al.  Support vector machine , 2013 .

[13]  Aamir Saeed Malik,et al.  Segmentation of satellite imagery based on pulse-coupled neural network , 2015, 2015 International Conference on Space Science and Communication (IconSpace).

[14]  Heggere S. Ranganath,et al.  Perfect image segmentation using pulse coupled neural networks , 1999, IEEE Trans. Neural Networks.