Infant growth obstacle intelligent evaluation method based on three-dimensional eye movement characteristic

The invention discloses an infant growth obstacle intelligent evaluation method based on a three-dimensional eye movement characteristic. The method mainly comprises the three steps of establishing adouble-sphere model of an eyeball optical axis, a binocular vision axis and a head-eye relative space relation, extracting the three-dimensional eye movement characteristic, and performing intelligentidentification of the infant growth obstacle. The double-sphere model estimates the optical axis according to the three-dimensional eyeball data, and obtains the relative relation between the binocular vision axis and eyes on the head. In three-dimensional eye movement characteristic extraction, a new eye movement parameter is obtained after three-dimensional modeling. In intelligent identification of the infant growth obstacle, an algorithm is applied for performing intelligent cleaning, storage, analysis, mining and displaying on mass eye movement data, and an association between a typicalparameter and a specific growth function obstacle is obtained, thereby determining the type of the obstacle. The method overcomes restriction in obtaining the eye movement characteristic; data miningtechnology and artificial intelligence technology are applied, thereby supplying a new technical tool for more objectively and comprehensively evaluating development of the infant cognition capability.