An Improved Skeleton Extraction Method of Retina Image based on Objective Refraction

摘要

Objective refraction is based on retina image processing to get patients' refractive power. Due to retinopathy and the different reflectivity of eye retina, the misdistribution of gray level of the retina image leads to the loss of partial details of the image by using the existing skeleton extraction method, thus affecting the accuracy of the measurement results. This paper proposes a dynamic gray threshold skeleton extraction method, which uses different gray threshold to extract image skeleton on different regions of the image. The gray threshold setting is only associated with the gray strength distribution of the related neighborhood image range, avoiding the loss of partial details of the image. Experiment results show that this method has the advantages of skeleton extraction of complete retina image features and fast computing speed, which can effectively improve the accuracy of eye refractive power measurement.