Space counting method of crucible bubbles based on video images

This paper proposes a crucible bubble space counting method based on image slices. A digital microscope is used to photograph the crucible transparent layers of different depths to construct a discrete three-dimensional space in the depth direction. The bubble image is enhanced, segmented, flooded and filled to obtain a binary image of the bubble on the two-dimensional plane. Aiming at the problem of low accuracy and long time-consuming traditional Hough circle detection algorithm, an improved random Hough transform bubble detection algorithm is proposed. The algorithm is used to obtain the center and radius parameters of the bubbles, and then the multi-layer bubbles in the three-dimensional space of the crucible transparent layer are counted. Compared with the traditional method, the single-frame bubble recognition rate of this method is more than 95%, and the detection time is reduced by 55% compared with the traditional algorithm. Compared with the actual crucible bubble count, the multi-layer bubble count has a correct rate of more than 95%.