Comparing image preprocessing techniques for detection of bubbles in leaks

This paper aims to present image analysis studies carried out to improve the detection of bubble flow rate in underwater application due petrol extraction or transportation leaks or natural exudation. It exemplifies and explains all the obtained results through the used techniques related to color space transformations, image histogram and filtering in the domain of the space and frequency. The achieved results are compared with previous techniques by counting, in binary image, the quantity of pixels defining bubble edges.