Multi-source remote sensing image big data classification system design in cloud computing environment

Abstract

Due to the problems of poor classification and time-consuming in traditional multi-source remote sensing image big data classification system, it cannot meet the standard requirements for image big data classification in related fields. To solve the above problems, the multi-source remote sensing image data classification system under cloud computing environment is optimised. Following the line string transmission protocol architecture, relevant information is processed, transformed and fused. Data are transported to the host through protocol transmission. Based on above principle, the system hardware and software are designed. Detailed, designing hardware system refers to designing image sensor interface and system processing interface. The design of the system software part can be divided into two parts, including the two-wire serial protocol formulation and the image big data classification algorithm that provides users with initialisation operations. At the same time, the image is sharpened and the pixels are improved. Experimental verification results show that the system has good processing effect and short time consumption.