Object recognition method, descriptor generating method for object recognition, and descriptor for object recognition

A method of generating a descriptor for an object recognition, the method comprising:

• obtaining a first feature component, being a difference of brightness values between a reference point and a neighbor point;
• a second feature component, being a hue value of the neighbor point;
• and a third feature component, being an angle between a vector from the reference point to the neighbor point and a vector from the reference point to a projection point of the neighbor point, with respect to a point cloud of a three dimensional image of an object;
• obtaining a fourth feature component, being an angle between a dominant vector of the point cloud and a vector from the reference point to the projection point of the neighbor point;
• and
• generating a descriptor configured for the object recognition using the first to the fourth feature components.

A descriptor for an object recognition, the descriptor comprising:

• a first feature component, being a difference of brightness values between a reference point and a neighbor point of a point cloud of a three dimensional (3D) image of an object;
• a second feature component, being a hue value of the neighbor point;
• a third feature component, being an angle between a vector from the reference point to the neighbor point and a vector from the reference point to a projection point of the neighbor point;
• and
• a fourth feature component, being an angle between a dominant vector of the point cloud and the vector from the reference point to the projection point of the neighbor point.