Method for saving image data of a camera in an accident data recorder of a vehicle
First Claim
Patent Images
1. A method of saving data in an accident data recorder of a vehicle, comprising the following method steps:
- a) obtaining, from a camera of the vehicle, camera image data of a scene of surroundings of the vehicle,b) performing object recognition on the camera image data to identify objects in the surroundings from the camera image data and to produce object data of the objects that have been identified,c) compressing the camera image data by a lossy compression method to produce compressed image data,d) saving saved data, including odometric data and position data of the vehicle as well as the compressed image data and the object data, in a predefined storage structure in a storage unit,e) overwriting the saved data in the storage unit after reaching a predefined volume of the saved data in the storage unit,f) statically saving the saved data in response to a trigger signal, andg) reading out the object data, the odometric data and the position data that has been statically saved, and performing synthetic image generation based on at least the object data, the odometric data and the position data that has been read out, to generate synthesized images of the identified objects in a reconstructed scenario representing the scene.
1 Assignment
0 Petitions
Accused Products
Abstract
A method of saving image data of a camera in an accident data recorder of a vehicle involves: a) performing object recognition to determine object data regarding objects detected from the camera image data; b) compressing the camera image data by a lossy compression method to form compressed image data; c) storing the object data and the compressed image data in a storage unit; d) overwriting the data in the storage unit after a predefined volume of data has been stored; and e) statically saving the object data and the compressed image data in response to a trigger signal.
3 Citations
14 Claims
-
1. A method of saving data in an accident data recorder of a vehicle, comprising the following method steps:
-
a) obtaining, from a camera of the vehicle, camera image data of a scene of surroundings of the vehicle, b) performing object recognition on the camera image data to identify objects in the surroundings from the camera image data and to produce object data of the objects that have been identified, c) compressing the camera image data by a lossy compression method to produce compressed image data, d) saving saved data, including odometric data and position data of the vehicle as well as the compressed image data and the object data, in a predefined storage structure in a storage unit, e) overwriting the saved data in the storage unit after reaching a predefined volume of the saved data in the storage unit, f) statically saving the saved data in response to a trigger signal, and g) reading out the object data, the odometric data and the position data that has been statically saved, and performing synthetic image generation based on at least the object data, the odometric data and the position data that has been read out, to generate synthesized images of the identified objects in a reconstructed scenario representing the scene. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method comprising the steps:
-
a) obtaining, from a camera of a vehicle, camera image data of a scene of surroundings of the vehicle; b) by performing object recognition on the camera image data, identifying objects in the surroundings from the camera image data and producing object data of the objects that have been identified; c) obtaining position data and odometric data of the vehicle; d) storing the object data, the position data and the odometric data; e) compressing the camera image data to produce compressed image data; f) storing the compressed image data; and g) reading out the object data, the position data and the odometric data that had been stored, and performing synthetic image generation based on at least the object data, the position data and the odometric data that has been read out, to generate synthesized images of the identified objects in a reconstructed scenario representing the scene. - View Dependent Claims (11, 12, 13, 14)
-
Specification