Positionable video data collecting and tissue retrieval method

Abstract

The invention discloses a positionable video data collecting and tissue retrieval method. GPS positioning information, video shooting directions and camera parameters are recorded during audio and video collecting to form a positionable video, and the relation between video frames and space information is described through MultiPoint space elements, wherein each point in MultiPoint is used for storing the positioning information, lens direction and video time stamp of a corresponding video frame; on the basis, in combination with shooting parameters, the two-dimensional visible range of the video frames is established; the visible range of the video frames is described through MultiPolygon, and then a positionable video space database is established; meanwhile, space indexes are established for MultiPolygon fields; finally, the efficient and precise space retrieval based on the positionable videos is realized under the GIS technological support. The video retrieval result is more precise, and the retrieval requirements of people are better met.

Description

technical field [0001] The invention relates to the fields of video data processing and geographic information science, in particular to a positioning-capable video data collection and organization retrieval method. Background technique [0002] Video, as a kind of public media, has become an indispensable information carrier in daily life, and its application fields are becoming more and more extensive. Facing the rapid growth of massive video data, how to adopt a fast and efficient indexing method is a key issue in the field of video data processing. Since the 1970s, video indexing has undergone two methods: text-based indexing and content-based indexing. Text indexing refers to the semantic hierarchy of video content (such as buildings, characters, events, etc.) by manually extracting key information from videos. Based on the description above, the video index is built with keywords, while the content-based index is based on the color, texture, shape and other characteri...

AI Technical Summary

Problems solved by technology

[0005] The disadvantages of the current positionable video data collection and retrieval methods are: (1) In terms of data collection, the above-mentioned positionable video collection method focuses on the problem of synchronous transmission and display of video and GPS position time. Constructing a spatial index to achieve efficient location-based video retrieval has few considerations; at the same time, video shooting parameter information, such as lens focal length, lens viewing angle, lens magnification, lens orientation and other parameters, is not fully considered, and only GPS positioning information is considered.

(2) In terms of data organization, the shooting scene of the locatable video actually reflects a linear area, but the current organization method simply abstracts the GPS trajectory corresponding to the shooting video into a line model, and uses the retrieval of lines instead of surfaces This method will cause only the video clips whose shooting positions are within the retrieval range to be retrieved, and some video clips whose shooting scenes are actually within the retrieval range are missed even though the shooting positions are outside the area.

Images