火焰特征是反映火工品燃烧性能的最重要、最直观的因素。通过比较几种火焰采集方法的优缺点，根据火工品火焰燃烧时间短、速度快和非接触性的特点，基于近年来迅速发展的数字图像处理技术建立了机器视觉系统。选取了先进的高精度大容量缓存彩色面阵CCD相机和彩色图像采集卡，进行遥控测量，得到了比较精确的序列火焰图像。并运用数字图像处理技术，根据火焰颜色和形状对火焰图像进行了初步分析，为进一步图像处理分析提供了有力的依据。

The flame feature is the most important and straight forward factor to reflect the burning properties of priming systems. Several flame collection methods were compared. Due to the short burning time, fast and noncontact characteristics of priming systems, a machine vision system was established based on the digital image processing technology developing rapidly in these years. An advanced high-precision mass-buffer-memory color planar-array CCD camera and a color frame grabber card were adopted for remote sensing. A series of flame images were obtained precisely, and the flame images were analyzed with the digital image processing technology based on the flame color and shape. The technology provides a reference for further image processing and analysis.