Abstract: Provided is an Aobject-based short range measurement method, a short range measurement device, a short range measurement system, and a storage medium. The short range measurement method includes: identifying a target object, and acquiring border information about an ROI of the target object; acquiring two groups of geometric constraint points of the target object with respect to a left-eye camera and a right-eye camera respectively; acquiring pixel coordinates of each geometric constraint point and a border pixel size corresponding to the border information, and calculating a monocular distance estimation value of the target object; acquiring an overall disparity of the two groups of geometric constraint points, and calculating a binocular distance estimation value of the target object in accordance with the overall disparity; and acquiring a final measurement value in accordance with the monocular distance estimation value and the binocular distance estimation value.

Abstract: The techniques described herein relate to constructing and using seed audiences. In an embodiment, a method includes loading, by a processing device, a user event sequence, the user event sequence including a plurality of user events and a plurality of corresponding conversions; generating, by the processing device, a plurality of conversion neighborhoods based on the user event sequence, a given conversion neighborhood in the plurality of conversion neighborhood including at least one conversion rule and a set of user events from the plurality of user events; annotating, by the processing device, each conversion neighborhood in the plurality of conversion neighborhoods with categorical labels; and generating, by the processing device, seed audiences for each conversion neighborhood, a given seed audience including a ranked list of user events for each conversion rule associated with the conversion neighborhood.

Abstract: The techniques described herein relate to constructing and using seed audiences. In an embodiment, a method includes loading, by a processing device, a user event sequence, the user event sequence including a plurality of user events and a plurality of corresponding conversions; generating, by the processing device, a plurality of conversion neighborhoods based on the user event sequence, a given conversion neighborhood in the plurality of conversion neighborhood including at least one conversion rule and a set of user events from the plurality of user events; annotating, by the processing device, each conversion neighborhood in the plurality of conversion neighborhoods with categorical labels; and generating, by the processing device, seed audiences for each conversion neighborhood, a given seed audience including a ranked list of user events for each conversion rule associated with the conversion neighborhood.

Abstract: An RGB-D fusion information-based obstacle target classification method includes: collecting an original image through a binocular camera within a target range, and acquiring a disparity map of the original image; collecting a color-calibrated RGB image through a reference camera of the binocular camera within the target range; acquiring an obstacle target through disparity clustering in accordance with the disparity map and the color-calibrated RGB image, and acquiring a target disparity map and a target RGB image of the obstacle target; calculating depth information about the obstacle target in accordance with the target disparity map; and acquiring a classification result of the obstacle target through RGB-D channel information fusion in accordance with the depth information and the target RGB image.

Abstract: A method and a system for extracting a dense disparity map based on multi-sensor fusion are provided. The method includes: obtaining a left-eye image and a right-eye image in a same road scenario, and point cloud information about the road scenario; generating an initial cost volume map set in accordance with the left-eye image, the right-eye image and the point cloud information; performing multidirectional cost aggregation in accordance with the point cloud information and the initial cost volume map set, and creating an energy function in accordance with the cost aggregation; and solving an optimum disparity for each pixel in the left-eye image in accordance with the energy function, so as to generate the dense disparity map.

Abstract: Provided is an Aobject-based short range measurement method, a short range measurement device, a short range measurement system, and a storage medium. The short range measurement method includes: identifying a target object, and acquiring border information about an ROI of the target object; acquiring two groups of geometric constraint points of the target object with respect to a left-eye camera and a right-eye camera respectively; acquiring pixel coordinates of each geometric constraint point and a border pixel size corresponding to the border information, and calculating a monocular distance estimation value of the target object; acquiring an overall disparity of the two groups of geometric constraint points, and calculating a binocular distance estimation value of the target object in accordance with the overall disparity; and acquiring a final measurement value in accordance with the monocular distance estimation value and the binocular distance estimation value.

Abstract: An automatic model reconstruction method and an automatic model reconstruction system for a component recognition model are provided. The automatic model reconstruction method includes the following steps. A first component image of a plurality of circuit boards is sequentially captured at a first position. The component recognition model sequentially recognizes component categories of the first component images, and a number of recognition probability values are output. According to the recognition probability values, a number of exponentially weighted moving averages (EWMA) are obtained. The first component images corresponding to the exponentially weighted moving averages lower than a first set value are collected until one of the exponentially weighted moving averages is lower than or equal to a second set value. The collected first component images are regarded as abnormal component images. The component recognition model is reconstructed according to the abnormal component images.

Abstract: An RGB-D fusion information-based obstacle target classification method includes: collecting an original image through a binocular camera within a target range, and acquiring a disparity map of the original image; collecting a color-calibrated RGB image through a reference camera of the binocular camera within the target range; acquiring an obstacle target through disparity clustering in accordance with the disparity map and the color-calibrated RGB image, and acquiring a target disparity map and a target RGB image of the obstacle target; calculating depth information about the obstacle target in accordance with the target disparity map; and acquiring a classification result of the obstacle target through RGB-D channel information fusion in accordance with the depth information and the target RGB image.

Abstract: One or more computing devices, systems, and/or methods for converting a data structure into an array are provided herein. Nodes of a data structure, such as a tree structure, are recursively processed to convert the data structure into an array. When processing a numerical node that is a parent of a low child node and a high child node, the numerical node of the tree structure is inserted into a first array element. The low child node is inserted into a second array element next to the first array element. The high child node is inserted into a third array element next to the second array element. A reference to the high child node is stored in association with the numerical node.

Abstract: Provided is a binocular camera depth calibration method, a binocular camera depth calibration device, a binocular camera depth calibration system and a storage medium. The method includes: acquiring a plurality of groups of images of a 2D code calibration board at different positions so as to acquire disparity maps of the 2D code calibration board; acquiring a 2D code template image of a target region; matching the 2D code template image, so as to determine and store matching region information; determining a position of the 2D code template image in each disparity map in accordance with the matching region information, and calculating an average disparity value of the 2D code template image at the position; acquiring a plurality of groups of average disparity values, and calculating a final disparity value; and calibrating a depth of a binocular camera.

Abstract: One or more computing devices, systems, and/or methods for converting a data structure into an array are provided herein. Nodes of a data structure, such as a tree structure, are recursively processed to convert the data structure into an array. When processing a numerical node that is a parent of a low child node and a high child node, the numerical node of the tree structure is inserted into a first array element. The low child node is inserted into a second array element next to the first array element. The high child node is inserted into a third array element next to the second array element. A reference to the high child node is stored in association with the numerical node.

Abstract: Provided is an object-based short range measurement method, a short range measurement device, a short range measurement system, and a storage medium. The short range measurement method includes: identifying a target object, and acquiring border information about an ROI of the target object; acquiring two groups of geometric constraint points of the target object with respect to a left-eye camera and a right-eye camera respectively; acquiring pixel coordinates of each geometric constraint point and a border pixel size corresponding to the border information, and calculating a monocular distance estimation value of the target object; acquiring an overall disparity of the two groups of geometric constraint points, and calculating a binocular distance estimation value of the target object in accordance with the overall disparity; and acquiring a final measurement value in accordance with the monocular distance estimation value and the binocular distance estimation value.

Abstract: Provided is a binocular camera depth calibration method, a binocular camera depth calibration device, a binocular camera depth calibration system and a storage medium. The method includes: acquiring a plurality of groups of images of a 2D code calibration board at different positions so as to acquire disparity maps of the 2D code calibration board; acquiring a 2D code template image of a target region; matching the 2D code template image, so as to determine and store matching region information; determining a position of the 2D code template image in each disparity map in accordance with the matching region information, and calculating an average disparity value of the 2D code template image at the position; acquiring a plurality of groups of average disparity values, and calculating a final disparity value; and calibrating a depth of a binocular camera.

Abstract: An image edge extraction device for an automated driving system is capable of extracting an image edge in a rapid manner. The device employs an image edge extraction method that includes: acquiring a grayscale image of a target image; calculating an edge magnitude and an edge angle of the grayscale image, so as to generate an edge magnitude image and an edge angle image; detecting an edge peak of the edge magnitude image in accordance with the edge angle image, so as to acquire each edge point of the edge magnitude image and acquire an edge peak detection result; and performing double-thresholding segmentation on the edge peak detection result through a flooding method, so as to acquire an image edge of the target image.

Abstract: A method for building a camera imaging model includes: converting world coordinate values of a random point P into camera coordinate values of a target camera in accordance with a predetermined mode; converting the camera coordinate values into image coordinate values of the target camera; and converting the image coordinate values into digital image coordinate values and building the camera imaging model. The converting the image coordinate values into the digital image coordinate values includes performing a spatial sampling operation on the random point P, and adjusting coordinate values of an origin to image coordinate values through calculation. The method can be used in an automated driving system for a vehicle.

Abstract: The present disclosure provides a detection method and a detection device for a digital camera. The detection method includes: Step 1 of moving the to-be-detected auto-focusing lens assembly to a position spaced apart from a subject by a first distance; Step 2 of performing automatic exposure control until an average luminance value of a central region of an image is in a stable state; Step 3 of locking an automatic exposure control parameter; Step 4 of taking a first image; Step 5 of determining sizes of four corner regions in the first image; Step 6 of counting average luminance values of the four corner regions; and Step 7 of, in the case that the average luminance values of the four corner regions are identical to each other, determining that the to-be-detected auto-focusing lens assembly is installed at an accurate position. According to the present disclosure, it is able to detect where or not the auto-focusing lens assembly is installed at the accurate position in an automatic and efficient manner.

Abstract: The present disclosure provides a method and a device for building a camera imaging model, and an automated driving system for a vehicle, so as to describe a camera imaging system through exterior and interior parameters of a camera acquired after calibration. The method includes: converting world coordinate values of a random point P into camera coordinate values of a target camera in accordance with a predetermined mode; converting the camera coordinate values into image coordinate values of the target camera; and converting the image coordinate values into digital image coordinate values and building the camera imaging model. The converting the image coordinate values into the digital image coordinate values includes performing a spatial sampling operation on the random point P, and adjusting coordinate values of an origin to image coordinate values through calculation.

Abstract: An image edge extraction device for an automated driving system is capable of extracting an image edge in a rapid manner. The device employs an image edge extraction method that includes: acquiring a grayscale image of a target image; calculating an edge magnitude and an edge angle of the grayscale image, so as to generate an edge magnitude image and an edge angle image; detecting an edge peak of the edge magnitude image in accordance with the edge angle image, so as to acquire each edge point of the edge magnitude image and acquire an edge peak detection result; and performing double-thresholding segmentation on the edge peak detection result through a flooding method, so as to acquire an image edge of the target image.

Abstract: The present disclosure provides a detection method and a detection device for a digital camera. The detection method includes: Step 1 of moving the to-be-detected auto-focusing lens assembly to a position spaced apart from a subject by a first distance; Step 2 of performing automatic exposure control until an average luminance value of a central region of an image is in a stable state; Step 3 of locking an automatic exposure control parameter; Step 4 of taking a first image; Step 5 of determining sizes of four corner regions in the first image; Step 6 of counting average luminance values of the four corner regions; and Step 7 of, in the case that the average luminance values of the four corner regions are identical to each other, determining that the to-be-detected auto-focusing lens assembly is installed at an accurate position. According to the present disclosure, it is able to detect where or not the auto-focusing lens assembly is installed at the accurate position in an automatic and efficient manner.

Abstract: A lane departure warning method includes steps of: collecting a road image by an imaging device; detecting a lane marking in accordance with the road image, so as to extract a position of the lane marking and an angle of the lane marking relative to a running direction of the vehicle; acquiring steering information and a movement speed of the vehicle by an OBD system; judging whether or not the vehicle is unconsciously running on the lane marking in accordance with the position of the lane marking, the angle of the lane marking relative to the running direction of the vehicle and the steering information of the vehicle; in the case that the vehicle is unconsciously running on the lane marking, recording a duration within which the vehicle is unconsciously running on the lane marking; and judging whether or not to send a lane departure warning to the vehicle.