Discussion on Multi sensor Data Fusion and Feature Extraction for Forestry Applications

This article is written for the data fusion by the multisensory as well as for extracting the feature which are used in the forestry applications. It also describes the levela of the feature at where the multi sensor using for data fusion according to the C-band, P- band and L-band in which also includes the data as the polar imetric synthetic aperture radar (PolSAR). This data is also includes as the Landsat of the multispectral according to the Thematic Mapper (TM).  In the ground of the France named as Nezer forest the application is classified for the class of the Maritime pine age class.

This articles is also discuss about the complementary information is used to motivation the data fusion of the multisensory. The optical data and SAR provides theses all complementary information. The major objective of these articles is to investigating the selection of the features between the well-known descriptors counted as 26. To improving the performance of the classification, befits of the multisensory data fusion are determined according g to the classification of the single sensor data with regards to the forest monitoring. By comparing the classification of the performances between the four dataset which is revealed that the best result are appearing as SAR feature and P band.

The improvement of the 12.6% is achieved as the accuracy of the classification by combining the multispectral optical features with SAR features and P-band. The four data sets have extracted the twenty six features whi9ch are instigated as the objective for identifying those features possessing jointly at maximum discrimination power. To preserving the classification of the 98.5% there are five features investigated this all information is compared with the set of the total features. Temesgen, Camilla and Anthony have written this article for showing the feature selection advantages according g to the preservation of the information of the classification at the same time for reducing the space of the feature. This articles is describes the potentials which is used for improving the performance of the classification it also can found by implementing the selection procedure of the feature (Temesgen Gebrie Yitayew, 2012).

Reference of Multi sensor Data Fusion and Feature Extraction for Forestry Applications