G06K 9/52 (20130101)
The present invention includes an automated method for grading the blurring of ocular fundus images to permit automated grading of severity of vitritis. Standard fundoscopic images are converted to gray-scale images. The gray-scale images are processed through a low-pass filter to eliminate high frequency components. The low pass image is subtracted from the corresponding original image, producing the difference image. The difference image is converted to the final entropy image. The entropy image is processed using Fast Fourier Transform to calculate its power. The final blur score is calculated.
Passaglia, Christopher L.; Madow, Brian; Richards, David W.; and Greenberg, Erin L., "Quantitative image analysis applied to the grading of vitreous haze" (2016). USF Patents. 884.
University of South Florida