Abstract
ABSTRACTThe understanding of mangrove forest structure and dynamics at species level is essential for mangrove conservation and management. To classify mangrove species, remote-sensing technologies provide a better way with high spatial resolution image. The spatial structure is usually viewed as effective complementary information for classification. However, it is still a challenge to design handcrafted features for mangrove species due to their non-structure texture. To leverage the advantage of convolutional neural networks (CNNs) in abstract feature exploration, a small patch-based CNN is proposed to overcome the requirement of fixed and large input which limits the applicability of CNNs to fringe mangrove forests. The function of down-sampling technology was substantially reduced to make deeper network for small input in our work. Meanwhile, the inception structure is used to exploit the multi-scale features of mangrove forests. Furthermore, the network is optimized with lesser convolution kernels and a single fully connected layer to reduce overfitting via reducing the training parameters. Compared to the features of grey level co-occurrence matrix with support vector machine, our proposed CNN shows better performance in classification accuracy. Moreover, the differences between mangrove species can be perceptive via CNN visualization, which offers better understanding of mangrove forests.
Highlights
Mangrove forests are the most productive ecosystems in the world with providing essential and unique ecological goods and services to human beings and adjacent systems including coastline and flood protection, shrimp farming and habitat provision for various terrestrial, estuarine and marine species (Giri et al 2011; Wang, Sousa, and Gong 2004; Polidoro et al 2010)
Since we focus on mangrove species differentiation, the main regions growing mangrove forests in both reserves (Figure 1) are depicted manually based on coarse classification using NDVI with a threshold of 0.2
This paper explores the ability of convolutional neural networks (CNNs) in mangrove species mapping and assesses the performance in feature extraction
Summary
Mangrove forests are the most productive ecosystems in the world with providing essential and unique ecological goods and services to human beings and adjacent systems including coastline and flood protection, shrimp farming and habitat provision for various terrestrial, estuarine and marine species (Giri et al 2011; Wang, Sousa, and Gong 2004; Polidoro et al 2010). Dominating the intertidal zone of tropical and subtropical coastlines, different mangrove species adapting to specific environment are distributed parallel to the coast or riverine system. They form distinct zones to be ‘foundation species’, which can control population and ecosystem dynamic, including fluxes of energy and nutrients, hydrology, food webs, and biodiversity (Polidoro et al 2010; Ellison et al 2005; Claridge and Burnett 1993). In terms of monitoring mangrove at species level, the median–low resolution remote-sensing images like Landsat TM are not appropriate due to coarser resolution (Wang, Sousa, and Gong 2004).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
