Multidirectional nystagmus signal extraction method based on nystagmus image

Abstract

The invention relates to an extracting and analyzing method for a horizontal nystagmus signal, a vertical nystagmus signal and a rotating nystagmus signal based on a nystagmus image. The method comprises the steps of decoding and normalizing; detecting and compensating reflection points; performing coarse positioning on an iris area; calculating a brightness gradient distribution matrix; constructing a shuttle-shaped ternary tree; calculating an optimal path; detecting the boundary of a pupil and the boundary of an iris; extracting the horizontal nystagmus signal and the vertical nystagmus signal; extracting the characteristic of the iris and coding and extracting the rotating nystagmus signal. According to the method, the nystagmus image is used as a processing object. Through video analysis and digital image processing technology, the nystagmus signals are extracted and visually displayed on a computer screen in a curve map manner, thereby supplying medically beneficial reference information.

Description

Technical field [0001] The present invention relates to the field of video image analysis and digital image processing, and in particular to a method of taking nystagmus images as objects and performing video decoding, image segmentation, feature extraction, feature encoding and template matching on them to obtain horizontal nystagmus images in nystagmus videos. Extraction and analysis methods for tremors, vertical nystagmus and rotational nystagmus. Background technique [0002] Nystagmus, also known as nystagmus, is a rhythmic movement of the eyeballs that is not controlled by thoughts. Vestibular nystagmus consists of alternating slow and fast phase movements. Nystagmus symptoms are closely related to central nervous system diseases, such as senile dementia, epilepsy, brain tumors, perivestibular lesions, central lesions (including brainstem and cerebellar lesions), vertigo and certain eye diseases. Therefore, the lesions of the above diseases can be understood by observ...

AI Technical Summary

Problems solved by technology

During the acquisition process, the three-dimensional eyeball information is projected onto the two-dimensional photosensitive element, resulting in interception and loss of information

Common problems are: A. The rotation of the eyeball causes the optical axis of the eyeball to be out of alignment with the optical axis of the camera, resulting in projection distortion; B. The off-axis distortion caused by the optical axis of the camera not being in line with the geometric center of the eyeball

(3) Very little research

Therefore, the interference factors proposed in (1, 2) can be avoided in the image, but it cannot be avoided in the nystagmus view detection

The Daugman calculus operator, Hough transform (Hough), active contour model (ActiveContour) and other methods used in the existing iris segmentation technology all perform poorly on complex eyeball images, resulting in system robustness. Poor, especially when distortion occurs, so it has not been widely used at present, and mainly depends on the subjective judgment of doctors

Images