The invention relates to an automatic detection method for microaneurysm in an eye
fundus image on the basis of a local entropy determining threshold. The method comprises steps as follows: step one, the image is preprocessed to realize uniformity of brightness of the eye
fundus image; step two,
Gaussian matched filtering is adopted to enhance pixels of blood vessels and the microaneurysm in the image; step three, the eye
fundus image is subjected to threshold segmentation by the aid of the local entropy determining threshold, and target areas of the blood vessels and the microaneurysm in the image are obtained through segmentation; step four, connected components extracted with
mathematical morphology are added after segmentation, length screening is realized, mutually communicating
blood vessel structures with the length exceeding a set threshold are removed from a selected dark area, and then a real microaneurysm structure can be detected; step five,
noise disturbance points in the image are removed through double-loop filtering, and the accuracy rate of detection is increased. Compared with the prior art, the method has the advantages of high detection accuracy and the like.