Method and apparatus for obtaining vehicle loss assessment image, server and terminal device

1. A computer-implemented method for obtaining a vehicle loss assessment image comprising:

receiving video data of a damaged vehicle;

detecting a plurality of video images in the video data to identify a damaged portion in the plurality of video images;

classifying one or more video images of the plurality of video images into each of one or more candidate image classification sets of the damaged portion based on the identified damaged portion, wherein the one or more candidate image classification sets comprise a close-up image set including one or more video images displaying the damaged portion and a component image set including one or more video images displaying a vehicle component to which the damaged portion belongs; and

selecting a vehicle loss assessment image from the one or more candidate image classification sets according to a screening condition,

wherein classifying the one or more video images into the close-up image set comprises:

sorting video images including the damaged portion in a descending order of areas of a same damaged portion in the video images; and

selecting, from the sorted video images, first one or more video images or one or more video images in each of which a ratio of an area of the damaged portion to an area of the corresponding video image is greater than a threshold ratio.

2. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 1 , wherein the threshold ratio is a first threshold ratio, and classifying one or more video images into the close-up image set further comprises:

in response to determining that a ratio of an area of the damaged portion to that of a video image including the damaged portion is greater than a second threshold ratio, classifying the video image into the close-up image set.

3. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 1 , wherein the threshold ratio is a first threshold ratio, and classifying one or more video images into the close-up image set further comprises:

in response to determining that a ratio of a horizontal coordinate span of the damaged portion to a length of a video image including the damaged portion is greater than a second threshold ratio, and/or a ratio of a longitudinal coordinate span of the damaged portion to a height of the video image including the damaged portion is greater than a third threshold ratio, classifying the video image into the close-up image set.

4. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 1 , further comprising:

in response to detecting that at least one of the close-up image set and the component image set of the damaged portion is empty, or the one or more video images in the close-up image set do not cover the entire damaged portion, generating a video recording prompt message; and

sending the video recording prompt message to a terminal device that obtains the video data.

5. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 1 , further comprising:

tracking the damaged portion in the video data in real time to determine a region of the damaged portion in the video images; and

in response to the damaged portion being out of a video image and subsequently re-entering a video image, tracking the damaged portion again to determine a new region of the damaged portion in the video image based on image feature data of the damaged portion.

6. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 5 , further comprising:

sending information of the region of the tracked damaged portion to a terminal device for the terminal device to display the region of the damaged portion in real time.

7. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 6 , further comprising:

receiving new information of the damaged portion, wherein the new information of the damaged portion is determined in response to the terminal device's changing the region of the damaged portion based on a received interactive instruction; and

classifying the video images based on the new information of the damaged portion.

8. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 1 , wherein selecting a vehicle loss assessment image from the one or more candidate image classification sets according to a screening condition comprises:

selecting at least one video image as a loss assessment image of the damaged portion from the one or more candidate image classification sets according to clarity of the video images and filming angles of the damaged portion in the video images.

9. The computer-implemented method for obtaining a vehicle loss assessment image according to claim 1 , further comprising:

in response to detecting that there are at least two damaged portions in the one or more video images, determining whether a distance between the at least two damaged portions satisfies a proximity condition; and

in response to determining that the distance between the at least two damaged portions satisfies the proximity condition, simultaneously tracking the at least two damaged portions, and obtaining loss assessment images of the at least two damaged portions.

10. An apparatus for obtaining a vehicle loss assessment image, comprising one or more processors and one or more non-transitory computer-readable memories coupled to the one or more processors and configured with instructions executable by the one or more processors to cause the apparatus to perform operations comprising:

receiving video data of a damaged vehicle;

detecting a plurality of video images in the video data to identify a damaged portion in the plurality of video images;

classifying one or more video images of the plurality of video images into each of one or more candidate image classification sets of the damaged portion based on the identified damaged portion, wherein the one or more candidate image classification sets comprise a close-up image set including one or more video images displaying the damaged portion and a component image set including one or more video images displaying a vehicle component to which the damaged portion belongs; and

selecting a vehicle loss assessment image from the one or more candidate image classification sets according to a screening condition,

wherein classifying the one or more video images into the close-up image set comprises:

sorting video images including the damaged portion in a descending order of areas of a same damaged portion in the video images; and

selecting, from the sorted video images, first one or more video images or one or more video images in each of which a ratio of an area of the damaged portion to an area of the corresponding video image is greater than a threshold ratio.

11. The apparatus for obtaining a vehicle loss assessment image according to claim 10 , wherein the threshold ratio is a first threshold ratio, and classifying one or more video images into the close-up image set further comprises:

in response to determining that a ratio of an area of the damaged portion to that of a video image including the damaged portion is greater than a second threshold ratio, classifying the video image into the close-up image set.

12. The apparatus for obtaining a vehicle loss assessment image according to claim 10 , wherein the threshold ratio is a first threshold ratio, and classifying one or more video images into the close-up image set further comprises:

in response to determining that a ratio of a horizontal coordinate span of the damaged portion to a length of a video image including the damaged portion is greater than a second threshold ratio, and/or a ratio of a longitudinal coordinate span of the damaged portion to a height of the video image including the damaged portion is greater than a third threshold ratio, classifying the video image into the close-up image set.

13. The apparatus for obtaining a vehicle loss assessment image according to claim 10 , wherein the operations further comprise:

in response to detecting that at least one of the close-up image set and the component image set of the damaged portion is empty, or the one or more video images in the close-up image set do not cover the entire damaged portion, generating a video recording prompt message; and

sending the video recording prompt message to a terminal device that obtains the video data.

14. The apparatus for obtaining a vehicle loss assessment image according to claim 10 , wherein the operations further comprise:

tracking the damaged portion in the video data in real time to determine a region of the damaged portion in the video images; and

in response to the damaged portion being out of a video image and subsequently re-entering a video image, tracking the damaged portion again to determine a new region of the damaged portion in the video image based on image feature data of the damaged portion.

15. The apparatus for obtaining a vehicle loss assessment image according to claim 14 , wherein the operations further comprise:

sending information of the region of the tracked damaged portion to a terminal device for the terminal device to display the region of the damaged portion in real time.

16. A non-transitory computer-readable storage medium configured with instructions executable by one or more processors to cause the one or more processors to perform operations comprising:

receiving video data of a damaged vehicle;

detecting a plurality of video images in the video data to identify a damaged portion in the plurality of video images;

classifying one or more video images of the plurality of video images into each of one or more candidate image classification sets of the damaged portion based on the identified damaged portion, wherein the one or more candidate image classification sets comprise a close-up image set including one or more video images displaying the damaged portion and a component image set including one or more video images displaying a vehicle component to which the damaged portion belongs; and

selecting a vehicle loss assessment image from the one or more candidate image classification sets according to a screening condition,

wherein classifying the one or more video images into the close-up image set comprises:

sorting video images including the damaged portion in a descending order of areas of a same damaged portion in the video images; and

selecting, from the sorted video images, first one or more video images or one or more video images in each of which a ratio of an area of the damaged portion to an area of the corresponding video image is greater than a threshold ratio.

17. The non-transitory computer-readable storage medium according to claim 16 , wherein the preset ratio is a first preset ratio, and classifying one or more video images into the close-up image set further comprises:

in response to determining that a ratio of an area of the damaged portion to that of a video image including the damaged portion is greater than a second preset ratio, classifying the video image into the close-up image set.

18. The non-transitory computer-readable storage medium according to claim 16 , wherein the preset ratio is a first preset ratio, and classifying one or more video images into the close-up image set further comprises:

in response to determining that a ratio of a horizontal coordinate span of the damaged portion to a length of a video image including the damaged portion is greater than a second preset ratio, and/or a ratio of a longitudinal coordinate span of the damaged portion to a height of the video image including the damaged portion is greater than a third preset ratio, classifying the video image into the close-up image set.

19. The non-transitory computer-readable storage medium according to claim 16 , wherein the operations further comprise:

in response to detecting that at least one of the close-up image set and the component image set of the damaged portion is empty, or the one or more video images in the close-up image set do not cover the entire damaged portion, generating a video recording prompt message; and

sending the video recording prompt message to a terminal device that obtains the video data.

20. The non-transitory computer-readable storage medium according to claim 16 , wherein the operations further comprise:

tracking the damaged portion in the video data in real time to determine a region of the damaged portion in the video images; and

in response to the damaged portion being out of a video image and subsequently re-entering a video image, tracking the damaged portion again to determine a new region of the damaged portion in the video image based on image feature data of the damaged portion.