FIELD: data processing.
SUBSTANCE: invention relates to a data processing method and a vision system for a robotic device. The effect is achieved due to the stages at which data are obtained from the source of navigation data and the installed set of vision sensors, namely "point clouds" containing three spatial coordinates and radial velocities of each point, depth maps, semantic maps and feature vectors that determine the level assessment optical visibility; the obtained data is pre-processed both by converting the obtained "point clouds" from the sensor coordinate system into a single static coordinate system, and by converting depth maps into a "point cloud" based on information about the calibration parameters of the stereo camera with simultaneous comparison of each point of the class corresponding to it; the generated "point clouds" are processed by two parallel pipelines, namely, the pipeline for detecting, tracking and predicting the trajectory of dynamic obstacles, as well as the pipeline for calculating cross-section maps; instantaneous maps of the profile patency of the surrounding space are calculated; the accumulated semantic maps of the profile permeability of the surrounding space are calculated; the trajectories of movement of dynamic objects are predicted; an array of points of the predictive trajectory of objects is formed and data is transmitted to external systems that solve the problems of automatic traffic control, both about stationary obstacles and about moving obstacles.
EFFECT: increasing the accuracy of detecting traffic obstacles in any light conditions, including when there are factors that reduce optical visibility in the environment (rain, snow, fog, smoke, dust, etc.).
10 cl, 3 dwg
Title | Year | Author | Number |
---|---|---|---|
METHOD FOR TRACKING OBJECTS AT THE RECOGNITION STAGE FOR UNMANNED VEHICLES | 2022 |
|
RU2798739C1 |
METHOD OF LOCALIZING AND CONSTRUCTING NAVIGATION MAPS OF MOBILE SERVICE ROBOT | 2020 |
|
RU2740229C1 |
TRAFFIC CONTROL SYSTEM FOR ELECTRIC UNMANNED VEHICLES | 2022 |
|
RU2789153C1 |
ON-BOARD POSITIONING DEVICE OF A RAIL VEHICLE | 2023 |
|
RU2799734C1 |
METHOD AND SYSTEM FOR CAPTURING OBJECTS USING ROBOTIC DEVICE | 2020 |
|
RU2745380C1 |
ROBOTIC TRANSPORT DEVICE FOR MOBILE ROBOTIC REPAIR AND DIAGNOSTIC COMPLEX | 2023 |
|
RU2806129C1 |
METHOD OF AUTOMATIC TREATMENT OF LARGE-SIZE THIN WALL PRODUCTS | 2018 |
|
RU2679859C1 |
AUTONOMOUS MOBILE MULTIFUNCTIONAL ROBOTIC COMPLEX | 2017 |
|
RU2722301C1 |
METHOD FOR BUILDING A ROUTE AND CONTROLLING THE MOVEMENT OF A MOBILE SERVICE ROBOT IN RETAIL PREMISES | 2021 |
|
RU2769710C1 |
METHOD FOR MUTUAL CALIBRATION OF A VIDEO CAMERA AND A FAN LASER ILLUMINATION APPARATUS | 2021 |
|
RU2788666C1 |
Authors
Dates
2022-10-31—Published
2021-12-22—Filed