Research On Measurement Of Digital Micro-Image For Sprayed Droplets

Size of sprayed droplets is one of important indexes to measure properties of sprayers. On the basis of digital image processing and pattern recognizing techniques, a quantitative auto-measurement system was developed to diagnose the digital microscopic image of sprayed droplets.The regions of sprayed droplets can not segment from background effectively, because the digital microscopic image of sprayed droplets is very different in illumination and has dirty pixels or image noise. First, the average-grey and median filter smoothing algorithms are used to eliminate the heterogeneity of background. The results show that medium value filter is an optimum method in dispel image noise. Edge detector and gray distributing information are used to extract droplets edges.The crucial procedure of contour detection for microscopic image of sprayed droplets and measurement of granularity is to segment droplets from background by producing binary images. The singular threshold is uneffective because the image is very different in illumination. In this paper, a fast maximum entropy image thresholding algorithm based on local gray feature is presented to exact sprayed droplets from background and preserve contours of droplets. Experiments show that this method can distinguish the object and the backgound perfectly and improve the binarization processing speed greatly.At last , The droplet sizing system can quantitatively auto-measure the dimension of sprayed droplets by scanning pixels. The experiments show that the measurement system is efficient, accurate and applied in the practice.