Vehicle and method of controlling the same

A vehicle comprising:

• a first camera configured to obtain a first image;
• a second camera configured to obtain a second image captured in a different field of view from the first camera;
• and
• a controller configured to
• obtain a distance between the vehicle and an object by processing images obtained by the first and second cameras,
• recognize a first object in a frame of the image,
• obtain a height of the first object, an aspect ratio of the first object, and a distance from the first object,
• assign the height, the aspect ratio, and the distance to 3D coordinate values,
• generate a 3D straight line based on a plurality of 3D coordinate values in each frame,
• store a characteristic relationship between the recognized first object and the 3D straight line in a memory,
• obtain a height of a second object and an aspect ratio of the second object by image processing, upon recognizing the second object of the same model as the first object through the multi-cameras, and
• estimate a distance from the second object based on the characteristic relationship,
• wherein the controller is configured to generate 3D coordinate values for each of a plurality of frames corresponding to the height, the aspect ratio, and the distance of the first object in a 3D Euclidean space, accumulate the generated 3D coordinate values, and fit the accumulated 3D coordinate values to the 3D straight line, and
• wherein the controller is configured to determine a necessity for automated online calibration (AOC) based on a first characteristic relationship for the first object and a second characteristic relationship for the second object, and process the AOC for the first image and second image based on the determined necessity for the AOC, and
• wherein the controller is configured to:
• determine the height of the object, which is corrected considering that the camera is turned in a yawing direction, the aspect ratio of the object, and the distance from the object, and
• input the height of the object, the aspect ratio of the object, and the distance from the object into a predetermined learning model to generate coordinate points corresponding to three vehicle features in the 3D Euclidean space and to fit the generated coordinate points to the 3D straight line.