System and method of acquiring coordinates of pupil center point

A method of acquiring coordinates of a pupil center point, comprising:

• acquiring a first image comprising a face and a second image comprising a pupil;
• extracting a first eye region corresponding to an eye from the first image;
• converting the first eye region into the second image, to acquire a second eye region corresponding to the eye in the second image;
• and
• detecting a pupil in the second eye region and acquiring the coordinates of the pupil center;
• wherein a resolution of the first image is smaller than a resolution of the second image,
• wherein the detecting the pupil and acquiring the coordinates of the pupil center point comprises:
• converting the second image in the second eye region into a grayscale image;
• calculating and acquiring an image gradient matrix of the grayscale image;
• calculating an adaptive threshold corresponding to the image gradient matrix;
• performing a normalization on each gradient value in the image gradient matrix in accordance with the adaptive threshold, to acquire an updated image gradient matrix;
• and
• calculating the coordinates of the pupil center point in accordance with the updated image gradient matrix,
• wherein the calculating the coordinates of the pupil center point in accordance with the updated image gradient matrix comprises:
• creating a weight matrix;
• selecting any point in the image gradient matrix as a target point calculating a sum of vector dot products between the target point and remaining points in the image gradient matrix in combination with the weight matrix;
• selecting a next target point and calculating a sum of vector dot products between the next target point and remaining points in the image gradient matrix until all points in the image gradient matrix are selected, to acquire a plurality of sums of vector dot products;
• and
• comparing all the sums of the vector dot products and taking coordinates of the target point corresponding to a maximum sum of vector dot products as the coordinates of the pupil center point;
• wherein the weight matrix is MAX-Px, y, MAX is a maximum grayscale value corresponding to a color bit depth, and Px, y is a pixel value in the grayscale image.

A system of calculating coordinates of a pupil center point, comprising a first camera, a second camera, a storage and a processor, wherein the first camera is configured to capture a first image comprising a face and output the first image to the processor, the second camera is configured to capture a second image comprising a pupil and output the second image to the processor, a resolution of the first camera is smaller than a resolution of the second camera, the storage is configured to store processing data, and the processor is configured to:

• acquire the first image and the second image;
• extract a first eye region corresponding to an eye from the first image;
• convert the first eye region into the second image, to acquire a second eye region corresponding to the eye in the second image;
• and
• detect a pupil in the second eye region and acquire the coordinates of the pupil center point,
• wherein the processor is configured to:
• convert the second image in the second eye region into a grayscale image;
• calculate and acquire an image gradient matrix of the grayscale image;
• and calculate the coordinates of the pupil center point in accordance with the image gradient matrix,
• wherein the processor is configured to:
• create a weight matrix;
• select any point in the image gradient matrix as a target point;
• calculate a sum of vector dot products between the target point and remaining points in the image gradient matrix in combination with the weight matrix;
• select a next target point and calculate a sum of vector dot products between the next target point and remaining points in the image gradient matrix until all points in the image gradient matrix are selected, to acquire a plurality of sums of vector dot products;
• and
• compare all the sums of the vector dot products and take coordinates of the target point corresponding to a maximum sum of vector dot products as the coordinates of the pupil center point;
• wherein the weight matrix is MAX-Px, y, MAX is a maximum grayscale value corresponding to a color bit depth, and Px, y is a pixel value in the grayscale image.