A wavelet method for analysis of droplet and particle images for monitoring heterogeneous processes

Image analysis has been used for many years in chemistry and chemical engineering laboratories for the study of size distributions and shapes of droplets/particles, and with the rapid progress in on-line digital imaging sensors, there is a great potential for applying the technique to on-line monitoring and automatic control of sizes and shapes of particulate products. One of the major challenges towards this goal is clearly the availability of methods for image analysis that need to be accurate, fast, robust, and tolerant of the quality of images and noises. This article describes a wavelet-based method for analysis of images obtained in heterogeneous polymerization. The method consists of four steps: image pre-processing using morphological operation, multi-scale wavelet analysis for edge detection, curvature-based circle recognition, and clustering. Real images from heterogeneous polymerization of varied qualities were used to illustrate the method.