A semiautomatic forest management approach using remote sensing techniques

Remote sensing satellite images are effectively used as a tool for decision making in various fields, especially in forest management and related analyses. Different geospatial parameters are required for effective decision making and the possibility of an integrated framework for automation of various analyses has been investigated. Advanced web mining and intelligent techniques have been adopted for the development of a comprehensive open-source framework for this purpose. The effectiveness of the developed methodology has been discussed and illustrated with reference to study areas using various statistical parameters. Adoption of a cellular neural network (CNN) for feature modeling and open-source data for automatic mining seemed to be effective. The developed methodologies were found to outperform existing ones with regard to accuracy and complexity. Investigations revealed that use of CNN is very effective in shape modeling, and improves accuracy of detection.