Abstract: Methods described herein relate to self-localization. Methods may include: receiving sensor data from a vehicle traveling along a road segment; identifying one or more objects of the environment from the sensor data; determining a coarse location of the vehicle; comparing the identified one or more objects with corresponding ground truth objects; calculating an observation in response to the comparison of the identified one or more objects with the corresponding ground truth objects; determining a location of the vehicle with a higher precision than the coarse location based, at least in part, on the observation; and allocating resources of the apparatus based, at least in part, on the observation.

Abstract: Apparatus and methods are described for roadway lane line detection. An aerial image including a plurality of pixels is received and provides to a classification model. The classification model provides probability values assigned to at least a portion of the plurality of pixels. A comparison of the probability values is performed to select at least one pixel according to the comparison. A lane line object is identified for the selected at least one pixel.

Abstract: An approach is provided for providing a digital elevation model. For example, the approach involves processing, by a processor, map elevation data to create a hierarchical resolution tile representation of the digital elevation model for a geographic area. The hierarchical resolution tile representation includes a plurality of resolution levels, wherein each of the plurality of levels represents the digital elevation model at a different resolution. The digital elevation model includes a plurality of control points, wherein each of the plurality of control points is associated with an elevation data point determined from the map elevation data.

Abstract: An approach is provided for providing a digital elevation model. For example, the approach involves processing, by a processor, map elevation data to create a hierarchical resolution tile representation of the digital elevation model for a geographic area. The hierarchical resolution tile representation includes a plurality of resolution levels, wherein each of the plurality of levels represents the digital elevation model at a different resolution. The digital elevation model includes a plurality of control points, wherein each of the plurality of control points is associated with an elevation data point determined from the map elevation data.

Abstract: Apparatus and methods are described for roadway lane line detection. An aerial image including a plurality of pixels is received and provides to a classification model. The classification model provides probability values assigned to at least a portion of the plurality of pixels. A comparison of the probability values is performed to select at least one pixel according to the comparison. A lane line object is identified for the selected at least one pixel.

Abstract: A method, apparatus and computer program product are provided to construct a digital elevation model utilizing ground points captured by ground-based light detection and ranging (LiDAR). In the context of a method, a plurality of ground points are extracted from a data set captured by ground-based LiDAR to construct a digital elevation model. Each ground point includes positional information including an altitude. The method also includes identifying one or more road structures that are altitudinally offset from surrounding terrain based upon the plurality of ground points and modifying the ground points of the digital elevation model that are representative of the one or more road structures that are altitudinally offset from the surrounding terrain. The method further includes at least partially filling regions in the digital elevation model that lack ground points captured by ground-based LiDAR.

Abstract: A method is provided for generating street-level imagery to provide unobstructed images, such as images of facades of a building, for use in on-line street viewing, three dimensional city rendering, research, or building based localization, among other uses. In particular, example methods may include obtaining a plurality of images of a first side of an object from a plurality of points-of-view, normalizing each of the plurality of images to obtain a plurality of normalized images of the first side of the object, where each of the plurality of normalized images includes a plurality of pixels, each with image location coordinates within a respective image. Method may further generate a merged image of the first side of the object, where the merged image is generated by comparing pixels between the plurality of normalized images with matching image location coordinates.

Abstract: Point cloud data is received and a ground plane is segmented. A two-dimensional image of the segmented ground plane is generated based on intensity values of the segmented ground plane. Lane marking candidates are determined based on intensity within the generated two-dimensional image. Image data is received and the generated two-dimensional image is registered with the received image data. Lane marking candidates of the received image data are determined based on the lane marking candidates of the registered two-dimensional image. Image patches are selected from the two-dimensional image and from the received image data based on the determined lane markings. Feature maps including selected image patches from the registered two-dimensional image and received data are generated. The set of feature maps are sub-sampled, and a feature vector is generated based on the set of feature maps. Lane markings are determined from the generated feature vector.

Abstract: A method, apparatus and computer program product are provided to construct a digital elevation model utilizing ground points captured by ground-based light detection and ranging (LiDAR). In the context of a method, a plurality of ground points are extracted from a data set captured by ground-based LiDAR to construct a digital elevation model. Each ground point includes positional information including an altitude. The method also includes identifying one or more road structures that are altitudinally offset from surrounding terrain based upon the plurality of ground points and modifying the ground points of the digital elevation model that are representative of the one or more road structures that are altitudinally offset from the surrounding terrain. The method further includes at least partially filling regions in the digital elevation model that lack ground points captured by ground-based LiDAR.

Abstract: Point cloud data is received and a ground plane is segmented. A two-dimensional image of the segmented ground plane is generated based on intensity values of the segmented ground plane. Lane marking candidates are determined based on intensity within the generated two-dimensional image. Image data is received and the generated two-dimensional image is registered with the received image data. Lane marking candidates of the received image data are determined based on the lane marking candidates of the registered two-dimensional image. Image patches are selected from the two-dimensional image and from the received image data based on the determined lane markings. Feature maps including selected image patches from the registered two-dimensional image and received data are generated. The set of feature maps are sub-sampled, and a feature vector is generated based on the set of feature maps. Lane markings are determined from the generated feature vector.

Abstract: A method, apparatus and computer program product are provided for determining a building location based on a building image. In the context of a method, the method includes receiving an image including at least one building, extracting building texture features from the image, and receiving reference building texture features. The method also includes determining, using a processor, building texture features comparisons between building texture features and reference building texture features which satisfy a predetermined match threshold and determining a matching building location based upon the building texture features comparison.

Abstract: A method, apparatus and computer program product are provided for roof type classification and reconstruction based on two dimensional aerial images. In the context of a method, the method includes receiving a roof image, determining a segmentation of the roof image based on cutting lines associated with roof features and classifying roof segments based on roof features within the segment. The classifying roof segments is based on the roof features correlation to a roof type pattern.

Abstract: A method is provided for generating street-level imagery to provide unobstructed images, such as images of facades of a building, for use in on-line street viewing, three dimensional city rendering, research, or building based localization, among other uses. In particular, example methods may include obtaining a plurality of images of a first side of an object from a plurality of points-of-view, normalizing each of the plurality of images to obtain a plurality of normalized images of the first side of the object, where each of the plurality of normalized images includes a plurality of pixels, each with image location coordinates within a respective image. Method may further generate a merged image of the first side of the object, where the merged image is generated by comparing pixels between the plurality of normalized images with matching image location coordinates.

Abstract: A method, apparatus and computer program product are provided for determining a building location based on a building image. In the context of a method, the method includes receiving an image including at least one building, extracting building texture features from the image, and receiving reference building texture features. The method also includes determining, using a processor, building texture features comparisons between building texture features and reference building texture features which satisfy a predetermined match threshold and determining a matching building location based upon the building texture features comparision.

Abstract: A method, apparatus and computer program product are provided for roof type classification and reconstruction based on two dimensional aerial images. In the context of a method, the method includes receiving a roof image, determining a segmentation of the roof image based on cutting lines associated with roof features and classifying roof segments based on roof features within the segment. The classifying roof segments is based on the roof features correlation to a roof type pattern.