Cloud detection in satellite images using multi-objective social spider optimization

摘要: Abstract Cloud detection algorithms have emerged to automate image data analysis because of its prime influential factor in remote sensing quality. algorithm still needs domain-expert intervention and large number training examples ensure good performance whose acquirement becomes difficult due unavailability labeled as well the time process heads involved. The paper puts forward multi-objective social spider optimization (MOSSO) based efficient clustering technique detect clouds visible range. This explains proposed MOSSO along-with carried on 14 benchmark two-objective test problems against MOEA/D, MODE, MOPSO SPEA2 algorithms. Further, strengths weaknesses are analyzed been used for implementation an named MOSSO-C. Optimal centroid matrix is attained MOSSO-C through environmental selection evaluation has done six synthetic databases compared with above mentioned conventional obtained results encourage use get neural network classifier. approach efficiently classifies cloudy pixels various Earth’s surfaces (water, vegetation land). also discusses four Landsat 8 which shows average 96.37% accuracy detecting pixels.