Abstract: It is the beneficial attempt to research the spray of fuel oil by using image process and analysis. Nevertheless, owning to the uneven brightness, the image of fuel droplet could not be segmented efficiently. Both homeostasis filtering and subtracting background are ordinary method adjust the uneven brightness. This paper analyzes these methods and concludes that subtracting background is superior to homeostasis filtering in adjusting the even brightness

Key words: Uneven brightness; Homeostasis filter; Subtract background; Liquid particle image

摘要： 运用数字图像处理与分析的方法来研究燃油的喷雾特性是一种有益的尝试。由于光照不均而造成的燃油喷雾液滴图像亮度不均，致使图像不能够有效进行阈值分割。采用同态滤波和减去背景算法是校正亮度不均的方法，通过比较分析，指出了在燃油喷雾液滴图像处理过程中减去背景算法是一种更加有效的校正亮度不均的方法。

关键词: 亮度不均；同态滤波；减去背景；液滴图像