国产欧美精品一区二区,中文字幕专区在线亚洲,国产精品美女网站在线观看,艾秋果冻传媒2021精品,在线免费一区二区,久久久久久青草大香综合精品,日韩美aaa特级毛片,欧美成人精品午夜免费影视

In view of the high-scoring earth observation system being affected by the constraints of different activities during the use process, the system imaging, backhaul and activity completion rate is low, resulting in poor observation results, for this reason, a convolutional neural network based on scale features is proposed. The design of the high-scoring Earth observation system. The system pushes system operating status information through the management and control center server to achieve the function of three-dimensional display tasks. The CMOS image sensor is used to realize the transmission of the corresponding points on the imaging surface, and the FPGA controller is used to control the data storage time. Adopt BCM5464 gigabit switch to realize high-speed data transmission. Construct and train the scale feature convolutional neural network, use the RPN network to identify the characteristics of the target area, and determine the training interest area coordinates in the area by dividing the foreground and background of the target, so that the RPN network weight learning achieves the expected goal and improves The accuracy of target detection and recognition, the design of the Earth observation information management process, and the completion of the system design. It can be seen from the experimental results that the highest imaging, return probability, and activity completion rate of the system are 83%, 99.9%, and 100%, respectively, which has good observation effects.