Image Processing And Volume Control For Droplet Evaporation Experiment

Droplet image processing and volume control system for space droplet evaporation experiment plays a significant role in monitoring and analyzing. However, due to the evaporation of droplet, we must continually inject liquid to maintain it during the experiment (around 10 hours); and because of its fluidity, contact methods could not be used to acquire the physical parameters, such as volume, surface area, surface tension, contact angle, etc. Droplet image processing and volume control system can solve these problems.The system consists of three parts: image acquisition, image processing and feedback control. First, the system captures an image through CCD camera, and then the image is compressed, transmitted and saved. After that, image preprocessing technology is used to get the physical parameters such as volume, surface area, surface tension, contact angle, etc. Finally, according to the present volume of the droplet, injector is controlled to inject liquid to keep the volume unchanged. The main work of this dissertation is:(1) Developed a lossless compression algorithm for droplet image sequence. First, some prevalent lossless compression algorithms are analyzed, such as lossless mode of JPEG, JPEG-LS, lossless mode of JPEG2000 and an algorithm based on integer wavelet transform and SPIHT. These algorithms are tested on 12 pictures from the standard test images, and the result shows that JPEG-LS is superior in compression ratio with a relative low computational complexity. In space droplet evaporation experiment, a great number of droplet images were captured, compressed and then transmitted to the base station. To improve lossless image compression ratio in space droplet evaporation experiment, according to the high correlation between droplet images that were sequentially captured, an image sequence lossless compression algorithm based on JPEG-LS algorithm and interframe prediction error coding was developed. The result of experiments shows that the proposed algorithm is superior to JPEG-LS in compression ratio and encoding time.(2) The application of edge detection algorithm in the image feedback control system. In order to get the physical parameters of the droplet, its contour must be detected first. There are several classic edge detectors, such as Robert edge detector, Sobel edge detector, Prewitt edge detector, LoG edge detector and Kirsch edge detector. Then Canny edge detector is introduced in details. By applying canny edge detection algorithm and then image thinning algorithm to the droplet image, a single pixel wide contour can be extracted. And experiments indicate that this contour is more accurate than that extracted by Sobel edge detector. (3) Application of PID control algorithm to control the volume of the droplet. The PID control theory and PID controller tuning is studied, and then a PID control algorithm is applied in the system. The PID control algorithm can control the injector and keep the volume of the droplet unchanged. This algorithm was used to control the volume of water droplet and alcohol droplet respectively, and the results proved its effectiveness.