Particle size measurement based on quality assessment of crystal images
Yixuan Yang, Yan Huo, Suisui Chen, Tao Liu · 2019
A synthetic image analysis method is proposed for in-situ particle size measurement used for monitoring crystallization processes, based on using a non-invasive imaging system. To facilitate crystal size measurement, a no-reference image quality assessment (NR-IQA) algorithm is proposed for crystal image analysis, based on image transformation, feature extraction, and a pre-trained support vector regression (SVR) model. Meanwhile, an adaptive thresholding algorithm is used to sieve out effective crystal images along the time evolution. Consequently, particle size measurement is conducted by computing the projection distances of primary axes identified from each crystal image. Experimental results on the cooling crystallization of $\beta$ form L-glutamic acid well demonstrate the effectiveness of the proposed image analysis method.