Abstract: A method for starting an application program in a Linux container and a corresponding Linux system are provided, where the method includes: creating, by a container start process, a container according to a command entered by a user, creating a first application start process, and putting the first application start process into the container for execution; obtaining, by the first application start process, according to the command entered by the user, the number N of application programs to be started, and creating N?1 second application start processes, where N is a positive integer not less than 2; and starting, in the container, by the first application start process and the N?1 second application start processes, the N application programs to be started. By using technical solutions of the present invention, when the container is created, a plurality of application programs can be started simultaneously.

Abstract: Systems, methods, and apparatus are described that that increase computer vision analysis in the field of semantic segmentation. With images accompanied by scan data, both two-dimensional and three-dimensional image information is employed for joint segmentation. Through the established correspondence between image data and scan data, two-dimensional and three-dimensional information respectively associated therewith is integrated. Using trained random forest classifiers, the probability of each pixel in images belonging to different object classes is predicted. With the predicted probability, optimization of the labeling of images and scan data is performed by integrating multiples cues in the markov random field.

Abstract: Systems, methods, and apparatus are described that that increase computer vision analysis in the field of semantic segmentation. With images accompanied by scan data, both two-dimensional and three-dimensional image information is employed for joint segmentation. Through the established correspondence between image data and scan data, two-dimensional and three-dimensional information respectively associated therewith is integrated. Using trained random forest classifiers, the probability of each pixel in images belonging to different object classes is predicted. With the predicted probability, optimization of the labeling of images and scan data is performed by integrating multiples cues in the markov random field.

Abstract: Systems and methods are provided to facilitate architectural modeling. In one aspect, a modeling system is provided. This includes a processor configured to generate remodeled images associated with one or more architectural structures. A procedural model is configured to integrate an approximate model from aerial images of the architectural structures and a three-dimensional (3D) reconstruction model from ground-level images of the architectural structures.