Method, apparatus, wearable helmet, image capturing apparatus and program for measuring distance based on image

1 . A method for measuring a distance using a processor, the method comprising:

acquiring a driving image of surroundings of a moving object;

acquiring tilt information according to driving of the moving object;

correcting the driving image using the tilt information; and

calculating a distance between the moving object and a target object included in the driving image based on the corrected driving image,

wherein the calculating the distance comprises calculating the distance between the moving object and the target object based on a height of a horizontal line in the driving image and a height of a bottom surface of a bounding box corresponding to the target object, and

wherein the method measures the distance between the moving object and the target object using the driving image captured in a driving environment of the moving object in which a rotation in a roll direction occurs.

2 . The method of claim 1 , wherein the tilt information is calculated based on a rotation value of a 3-axis sensor in a roll direction.

3 . The method of claim 2 , wherein the 3-axis sensor is provided in an image capturing apparatus for capturing the driving image of the moving object.

4 . The method of claim 3 , wherein

the image capturing apparatus includes:

a first image capturing apparatus capturing an image of a front of the moving object; and

at least one second image capturing apparatus capturing an image other than the front.

5 . The method of claim 1 , further comprising:

recognizing a target object in the driving image using a neural network model,

wherein the neural network model is a model trained to recognize an object in the driving image and output a bounding box representing an object region in the driving image.

6 . The method of claim 5 , wherein

the correcting includes rotating the driving image based on the tilt information, and

the recognizing includes recognizing a target object based on the rotated driving image.

7 . The method of claim 5 , wherein the correcting includes rotating a bounding box of a target object recognized from the driving image based on the tilt information.

8 . A non-transitory computer-readable recording medium in which a program for performing a method for measuring a distance according to claim 1 is stored.

9 . An apparatus for measuring a distance, the apparatus comprising:

a driving image acquiring unit acquiring a driving image of surroundings of a moving object;

a processor acquiring tilt information according to driving of the moving object;

the processor correcting the driving image using the tilt information; and

the processor calculating the distance between the moving object and a target object included in the driving image based on a corrected driving image,

wherein the processor calculates the distance comprises calculating the distance between the moving object and the target object based on a height of a horizontal line in the driving image and a height of a bottom surface of a bounding box corresponding to the target object, and

where in the apparatus is configured to measure the distance between the moving object and the target object using the driving image captured in a driving environment of the moving object in which a rotation in a roll direction occurs.

10 . The apparatus of claim 9 , wherein the tilt information is calculated based on a rotation value of a 3-axis sensor in a roll direction.

11 . The apparatus of claim 10 , wherein the 3-axis sensor is provided in an image capturing apparatus capturing a driving image of the moving object.

12 . The apparatus of claim 11 , wherein

the image capturing apparatus includes:

a first image capturing apparatus capturing an image of a front of the moving object; and

at least one second image capturing apparatus capturing an image other than the front.

13 . The apparatus of claim 10 , further comprising:

the processor recognizing a target object in the driving image using a neural network model,

wherein the neural network model is a model trained to recognize an object in the driving image and output a bounding box representing an object region in the driving image.

14 . The apparatus of claim 13 , wherein the processor rotates the driving image based on the tilt information, and the processor recognizes the target object based on the rotated driving image.

15 . The apparatus of claim 14 , wherein the processor rotates the bounding box of the target object recognized in the driving image based on the tilt information.

16 . A wearable helmet for a personal mobility device driver, the wearable helmet comprising:

a head up display (HUD) outputting guidance information that can be checked by a driver;

a driving image acquiring unit acquiring a driving image of surroundings of a moving object;

a processor acquiring tilt information according to driving of the moving object;

the processor correcting the driving image using the tilt information; and

the processor calculating a distance between the moving object and a target object included in the driving image based on the corrected driving image,

wherein the processor calculates the distance comprises calculating the distance between the moving object and the target object based on a height of a horizontal line in the driving image and a height of a bottom surface of a bounding box corresponding to the target object, and

wherein the wearable helmet is configured to measure the distance between the moving object and the target object using the driving image captured in a driving environment of the moving object in which a rotation in a roll direction occurs.

17 . The wearable helmet of claim 16 , further comprising:

a controller controlling the HUD to display a calculated distance.

18 . An image capturing apparatus installed in a personal mobility device, the image capturing apparatus comprising:

an image capturing unit capturing a driving image of surroundings of a moving object;

a processor acquiring tilt information according to driving of the moving object;

the processor correcting the driving image using the tilt information; and

the processor calculating a distance between the moving object and a target object included in the driving image based on the corrected driving image,

wherein the processor calculates the distance comprises calculating the distance between the moving object and the target object based on a height of a horizontal line in the driving image and a height of a bottom surface of a bounding box corresponding to the target object, and

wherein the image capturing apparatus is configured to measure the distance between the moving object and the target object using the driving image captured in a driving environment of the moving object in which a rotation in a roll direction occurs.

19 . The image capturing apparatus of claim 18 , further comprising:

the processor transmitting data on the distance between the moving object and the target object to a wearable helmet for a driver of the personal mobility device.