Abstract: A method and an apparatus for generating an offline map are disclosed. The method includes: obtaining 1st to nth environmental images sequentially collected by a camera of a vehicle when the vehicle is traveling on a target road section; selecting ith to jth environmental images from the 1st to Nth environmental images, performing a first batch of three-dimensional modeling on the ith to jth environmental images to obtain first posture information of the camera and a first world coordinate; performing a second batch of three-dimensional modeling in parallel on the 1st to (i?1)th environmental images and (j+1)th to Nth environmental images, based on the first posture information and the first world coordinate, to obtain a second world coordinate and to obtain a third world coordinate; and generating an offline map of the target road section according to the world coordinate.

Abstract: The present disclosure discloses a method for generating a parking model, an electronic device and a storage medium, and relates a field of autonomous parking technologies. The detailed implementing solution includes: obtaining multiple sample sets; constructing a parking cruise space for the target vehicle based on each sample set, and extracting a first parking trajectory corresponding to each sample set from each parking cruise space; recognizing an abnormal position on each first parking trajectory, and deleting driving data corresponding to the abnormal position from a sample set corresponding to each first parking trajectory to obtain target sample data in each sample set; and performing model training based on the target sample data in each sample set to generate a target parking model.

Abstract: The present application discloses a verification method and device for a modeling route, an unmanned vehicle, and a storage medium, which relate to the technical field of computer vision and intelligent transportation. A specific implementation of the method in the present application lies in: acquiring a filtering threshold of a target road section, where the filtering threshold is related to image feature points corresponding to the target road section; verifying a modeling route corresponding to the target road section through the filtering threshold to obtain a verification result. According to the present application, availability of the modeling route can be directly verified with the filtering threshold while there is no need to verify the modeling route through manual driving of the vehicle, thereby effectively increasing the verification efficiency, protecting the vehicle from travelling along an unavailable modeling route and improving the driving experience.

Abstract: The present disclosure discloses a method for generating a parking model, an electronic device and a storage medium, and relates a field of autonomous parking technologies. The detailed implementing solution includes: obtaining multiple sample sets; constructing a parking cruise space for the target vehicle based on each sample set, and extracting a first parking trajectory corresponding to each sample set from each parking cruise space; recognizing an abnormal position on each first parking trajectory, and deleting driving data corresponding to the abnormal position from a sample set corresponding to each first parking trajectory to obtain target sample data in each sample set; and performing model training based on the target sample data in each sample set to generate a target parking model.

Abstract: A method and an apparatus for generating an offline map are disclosed. The method includes: obtaining 1st to Nth environmental images sequentially collected by a camera of a vehicle when the vehicle is traveling on a target road section; selecting ith to jth environmental images from the 1st to Nth environmental images, performing a first batch of three-dimensional modeling on the ith to jth environmental images to obtain first posture information of the camera and a first world coordinate; performing a second batch of three-dimensional modeling in parallel on the 1st to (i?1)th environmental images and (j+1)th to Nth environmental images, based on the first posture information and the first world coordinate, to obtain a second world coordinate and to obtain a third world coordinate; and generating an offline map of the target road section according to the world coordinate.

Abstract: The present application discloses a verification method and device for a modeling route, an unmanned vehicle, and a storage medium, which relate to the technical field of computer vision and intelligent transportation. A specific implementation of the method in the present application lies in: acquiring a filtering threshold of a target road section, where the filtering threshold is related to image feature points corresponding to the target road section; verifying a modeling route corresponding to the target road section through the filtering threshold to obtain a verification result. According to the present application, availability of the modeling route can be directly verified with the filtering threshold while there is no need to verify the modeling route through manual driving of the vehicle, thereby effectively increasing the verification efficiency, protecting the vehicle from travelling along an unavailable modeling route and improving the driving experience.

Abstract: A method and an apparatus for image processing for a vehicle are provided. The method includes: acquiring an image presenting a target object; occluding at least one area of the image to obtain at least one occluded image; inputting the at least one occluded image into a pre-trained image processing model to obtain a driving parameter corresponding to each of the at least one occluded image; determining, for each driving parameter, a difference between the driving parameter and a real driving parameter corresponding to the acquired image; and determining whether the difference between the driving parameter of the occluded image occluding at least one area where the target object is located and the real driving parameter is greater than or equal to a difference threshold, to determine a degree of association between the state of the target object and the driving parameter.

Abstract: The present invention provides an FPGA based data acquisition card, data acquisition system and data acquisition method. The FPGA based data acquisition card comprises: a plurality of bus control modules and a plurality of sensor acquisition modules constructed on an FPGA on the basis of a soft core, the plurality of bus control modules being used to acquire bus data of each of a plurality of bus devices, the plurality of sensor acquisition modules being used to acquire sensor data on a plurality of sensors; and a controller on the FPGA, the controller being used to send the acquired bus data and sensor data to a processing device connected to the data acquisition card.

Abstract: A method and an apparatus for image processing for a vehicle are provided. The method includes: acquiring an image presenting a target object; occluding at least one area of the image to obtain at least one occluded image; inputting the at least one occluded image into a pre-trained image processing model to obtain a driving parameter corresponding to each of the at least one occluded image; determining, for each driving parameter, a difference between the driving parameter and a real driving parameter corresponding to the acquired image; and determining whether the difference between the driving parameter of the occluded image occluding at least one area where the target object is located and the real driving parameter is greater than or equal to a difference threshold, to determine a degree of association between the state of the target object and the driving parameter.

Abstract: A data training method and a data training apparatus for an autonomous vehicle are provided according to embodiments of the present disclosure. The method includes: acquiring sensor data of the autonomous driving vehicle and a correction sample, the correction sample being used for representing driving behavior data of a driver when the autonomous driving vehicle encounters an interference during driving; and building an end-to-end model using the sensor data and the correction sample, the end-to-end model being configured for outputting a control instruction corresponding to a driving behavior of the driver using the sensor data and the correction sample. According to the embodiments, driving safety of the autonomous vehicle is improved.