油田井场中的火灾监控是油井安全生产的一个重要方面。基于红外传感器和烟雾传感器的传统的火灾报警装置 ,并不完全适用于野外井场等大空间场合的火灾监控。针对井场等大空间场合的火灾监控的特点 ,文章提出了一个基于火灾燃烧时火焰红外图象序列的特征识别的火灾自动识别监控方法 ,并使用该方法构成了一个用于井场防火监控的多通道火灾实时自动报警系统样机。实验结果表明 ,该火灾自动报警系统对于一般油田井场等大空间场合的火灾监控报警是有效的 Fire monitoring in oilfields is an important aspect of well safety. Traditional fire alarm devices based on infrared sensors and smoke sensors are not entirely suitable for fire monitoring in large space applications such as field well sites. In view of the characteristics of fire monitoring in the case of large space such as well site, this paper presents a fire automatic identification and monitoring method based on the feature recognition of flame infrared image sequence when fire is burnt, and uses this method to form a fire monitoring system Multi-channel fire real-time automatic alarm system prototype. The experimental results show that the automatic fire alarm system is effective for fire monitoring and alarming in the case of large oil field such as well field