Analysis on application of swarm-based techniques in processing remote sensed data

Snehlata Sheoran, Neetu Mittal, Alexander Gelbukh

    Research output: Contribution to journalArticlepeer-review

    1 Scopus citations

    Abstract

    The remote sensed satellite images are big repository of information and provide the coverage of large areas. However, these images may not be able to describe the finer details of area being covered. Satellite Image optimization is the process of augmenting the components of an image for better and effective interpretations from satellite images. In order to obtain better visibility properties to fetch more information, various artificial intelligence techniques can be considered for the optimization process. Finding out the best technique for optimization is a challenging and time-consuming task [U1]. In this paper, applications of swarm-based artificial intelligence techniques such as ant colony optimization, particle swarm optimization, bat algorithm, artificial bee colony algorithm etc. are being analysed to process the remote sensed data. The detailed comparison with respect to classifier, utility, images considered, and observation are discussed. The comprehensive analysis revealed that particle swarm optimization is the most widely used technique. Further, various application areas such as land-use land-cover are discussed with possibilities of future research [U2].

    Original languageEnglish
    Pages (from-to)97-113
    Number of pages17
    JournalEarth Science Informatics
    Volume13
    Issue number1
    DOIs
    StateAccepted/In press - 1 Jan 2019

    Keywords

    • Artificial intelligence (AI)
    • Change detection
    • Classification
    • Image processing
    • Land-use land-cover
    • Optimization
    • Particle swarm optimization (PSO) [U3]
    • Remote sensing
    • Satellite images
    • Segmentation
    • Swarm intelligence

    Fingerprint Dive into the research topics of 'Analysis on application of swarm-based techniques in processing remote sensed data'. Together they form a unique fingerprint.

    Cite this