| 摘要 |
When a hand of the user is recognized in an image pickup region of a camera, a head mounted display stores a contour shape of the hand which would be imaged by the camera in advance. In addition, the head mounted display receives an input of image data per pixel included in the camera, calculates a difference between colors of adjacent pixels represented by the image data, sets a set of image data having the same color system, where the calculated difference is within a predetermined threshold, as a group, and captures a contour of a region of the data. Next, the head mounted display compares the captured contour to a contour shape of the hand which is stored in advance to allow the user to recognize the hand of the user in the image pickup region. |
| 主权项 |
1. An image processing device for recognizing a hand of user in an image pickup region of a camera which picks up an image of a front side of the user, the device comprising:
a difference calculating unit that receives an input of a picked-up image composed of a plurality of data included in the camera, calculates a difference between the data of the picked-up image, and executes the difference calculation between the data in at least a part of the image pickup region, wherein the data of the picked-up image is color-converted into HSV color-based data in which color is re resented by hue, saturation and value, and the difference calculation is between adjacent pixels and is a calculation of a difference between values of adjacent pixels obtained by the color conversion, and the value obtained by the color conversion of the data of the picked-up image is separated from a brightness affected by the illuminance around an imaged hand of a user during the color conversion into an HSV color system; a capturing unit that captures a shape of a region, where the difference calculated by the difference calculating unit is within a predetermined threshold, as a recognition target shape; a storing unit in which a contour shape of the hand is stored in advance as a comparative shape of the captured recognition target shape; an unevenness comparison unit that compares captured contours to the contour shape of the hand being stored in the storage unit, the contour shape of the hand being stored in advance and being a predetermined shape, the comparison occurring after the capturing of the contours, wherein the contour shape of the hand being stored in the storage unit has data representing a plurality of configurations relating to an orientation of the hand; and a recognizing unit that compares the recognition target shape captured by the capturing unit to the contour shape stored in the storing unit and recognizes the captured recognition target shape as the hand of the user in the image pickup region based on a comparison result, wherein a pointer is provided based upon converted coordinates when the image pickup region is modified based upon a virtual image, and displayed at converted coordinates of the virtual image so as to correspond to a point on a user's hand. |
