| 摘要 |
An image processing unit determines the size of a circumscribed quadrangle of the target object, and predicts the overlap of the two different objects. If overlap is predicted, the target object is divided into two divided regions L_b and R_b containing the left and right edges of the target object respectively, and by performing a correlation calculation between the left image and the right image captured by a stereo camera for each of the divided regions, the parallaxes of the respective divided regions L_b and R_b can be determined. Then, if the absolute value of the difference between the parallaxes of the respective divided regions L_b and R_b is greater than a predetermined value, a determination is made that different objects are overlapping within the target object, and the target object is divided for recognition, and the distance to the respective objects is calculated independently.
|
