现有军用光电装备快速智能化手段

近年来,人工智能技术在图像处理领域取得了突破性进展,使得目标识别和目标检测算法的性能得到大幅提高,也为军用光电装备智能化创造了有利条件。本文从算法、数据和计算能力等人工智能三要素入手,分析了快速实现军用光电装备智能化的有效途径及过程中可能面临的问题,指出:现有的技术理论基本满足装备升级需求,智能化过程中还应重点加强实时性算法的研究,提早对专用数据集的收集整理进行布局完善,采用低功耗高性能边缘计算设备为装备计算能力水平提升提供支撑等。同时,在工程化过程中仍存在部分实际问题,需根据装备进行个性化设计,以解决理论与应用之间的差距。最后,建议重点加强国产化水平与能力的提高,以实现关键技术自主可控。

Rapid intelligentization of military optoelectronic equipment

In recent years, Artificial Intelligence technology has made remarkable breakthroughs in the field of image processing to significantly improve the performance of object recognition and detection algorithms, where this has created favorable conditions for the intelligentization of military optoelectronic equipment. The article begins by elucidating the three elements of Artificial Intelligence: algorithms, data, and computing capabilities. It then analyzes effective ways to quickly realize the intelligentization of military optoelectronic equipment and the problems that may be encountered in the process. The discussion shows that current technical theories meet the requirements of equipment upgrades. In the process of intelligentization, such issues as real-time algorithms, special datasets, and the deployment of low-power high-performance edge computing devices need to be considered. At the same time, practical problems persist in the relevant engineering such that personalized design is needed to bridge the gap between theory and application. Finally, the authors propose focusing on strengthening the standards and capabilities of localization for the independent control of key technologies.