微小型无人机载光电吊舱显控软件的设计与实现
2022,30(3):145-150
摘要：针对微小型光电吊舱的特点，使用 Qt框架和C/C++语言设计并实现了某型机载微小型光电吊舱的显示与控制软件；该软件在设计上采用了多线程的方式，分别实现了基于QtSerialPort模块的串口数据收发；基于开源编解码库FFmpeg的视频流接收和解码；基于YOLOv3模型检测算法的目标检测和识别等多项任务；各模块内部独立使用线程，并利用Qt的信号槽机制传递消息，使软件程序层次清晰，可读性强，并有利于后续功能的修改和添加；在地面高性能设备中使用YOLOv3算法做目标识别，在视频延时符合要求时，明显提升了检测速度和识别精度；最后，通过多次实验，验证了该显控软件的可靠性和有效性。
关键词：光电吊舱；显控软件；Qt；目标识别；多线程；C++； YOLOv3
Design and implementation of display-console software for micro unmanned electro-optical pod
Abstract：According to the characteristics of micro electro-optical pod, using the Qt framework and C/C++ language to design and implement the display and control software of a certain type of airborne micro-photoelectric pod; the software is designed in a multi-threaded way to realize the serial ports based on the QtSerialPort module. Data transmission and reception; video stream reception and decoding based on the open source codec library FFmpeg; multiple tasks such as target detection and recognition based on the YOLOv3 model detection algorithm; each module uses threads independently, and uses Qt's signal slot mechanism to transmit messages to make the software The program level is clear, the readability is strong, and it is conducive to the modification and addition of subsequent functions; the YOLOv3 algorithm is used for target recognition in the ground high-performance equipment, and the detection speed and recognition accuracy are significantly improved when the video delay meets the requirements; finally , after many experiments, the reliability and effectiveness of the display and control software have been verified.
Key words：electro-optical pod; display console software; Qt; target recognition; multithreading; C++; YOLOv3
收稿日期：2021-12-22
基金项目：
