All-round stereoscopic vision matching system and method based on neural network and medium
A technology of stereo vision matching and neural network, which is applied in neural learning methods, biological neural network models, image data processing, etc., can solve problems that cannot meet the needs of unmanned system application scenarios, and achieve easy large-scale promotion and low cost Effect
Pending Publication Date: 2021-04-09
SUN YAT SEN UNIV
View PDF7 Cites 6 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, the traditional stereo vision system is based on binocular vision, and binocular vision can only obtain depth information in a single direction. Information cannot be used to make decisions such as autonomous obstacle avoidance and autonomous navigation, and a single depth of information can no longer meet the application scenario requirements of unmanned systems
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0053] Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.
[0054] In the description of the present invention, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc. indicated orientations or positional relationships are based on the orientations or positional relationships shown in the drawings, and are only In order to facilitate the description of the present invention and simplify the description, it does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an all-round stereoscopic vision matching system and method based on a neural network and a storage medium. The system comprises a feature extraction module, a stereoscopic matching module, a processing module and a splicing module. Through looking around depth information in multiple directions, the system can be well used for automatic obstacle avoidance and autonomous navigation decision of the unmanned system, improves the degree of freedom of unmanned systems such as unmanned aerial vehicles and unmanned vehicles, and expands the application scenarios of the unmanned systems; the whole system adopts a hardware-friendly design and can be easily deployed in a hardware system, and the processing speed of the system is improved. Besides, the neural network is used for feature extraction, the advantage of high precision is achieved, and the system can be applied to application scenes with high real-time performance and high precision requirements. And on the other hand, compared with laser radar, structured light, TOF and other technologies, the system provided by the embodiment of the invention is low in cost and easy to popularize on a large scale. The system and the method can be widely applied to the technical field of image processing.
Description
technical field [0001] The invention relates to the technical field of image processing, in particular to a neural network-based surround-view stereo vision matching system, method and storage medium. Background technique [0002] Depth information is one of the most important information in the field of computer vision. Currently, the technologies used to obtain depth information mainly include lidar, structured light, TOF and stereo vision. Among these technologies, lidar can obtain the most accurate depth information, but lidar is very expensive, the depth is sparse, there is no visual information, and it is large in size, heavy in weight, and high in power consumption, so it cannot be deployed to drones, robots, etc. For unmanned systems; structured light and TOF technology are easily affected by external light when used indoors, but when used outdoors, the accuracy is very low, while the distance is short and the cost is relatively high. [0003] Although the stereo vi...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06K9/46G06K9/38G06T3/40G06T5/00G06N3/08
CPCG06T3/4038G06N3/08G06T2207/20032G06V10/28G06V10/40G06F18/22G06T5/70
Inventor 陈刚何涛张余凌晔华黄凯
Owner SUN YAT SEN UNIV