Research on Image Recognition Algorithm for Intraocular Pressure Monitoring at Home

In order to realize remote online continuous monitoring of intraocular pressure fluctuation data at home, through image processing, feature extraction and SVM classification algorithm, the intraocular pressure fluctuation data displayed on the screen of the self-developed wearable intraocular pressure monitoring terminal was identified. Smooth the image through image filtering technology. Image segmentation technology prepares three types of intraocular pressure data image sample sets. The HOG feature extraction algorithm extracts their gradient size and direction. Using the SVM classification algorithm, three types of intraocular pressure gradient data are classified to obtain a training model. Each time the intraocular pressure is detected, the final intraocular pressure fluctuation data is obtained by comparing the training model.