Technology

Spatially Encoded Video

Applications

Benefits

Introduction

Spatially encoded video is analogue or digital video which has  been tagged with some sort of real world reference such as X, Y coordinates or street address. This ‘georeferenced’ video can then be linked to other geographical datasets inside GI systems and in turn used to provide additional data about the world around us. The central challenge spatial video integrators face is how to integrate this rich, dynamic, spatial content with conventional GI datasets. Video, Navigation, Photogrammetry and GIS technologies are required to integrate these various geo datasets, which possess widely varying spatial, spectral, temporal and metric attributes.

Spatial multimedia systems not only automate mapping tasks but also increase the range of roles which can now be carried out within the GI sector. These include a large number of  asset inventory and infrastructure mapping tasks previously carried out manually, for example, rail and pipeline mapping. Spatially monitoring and mapping systems are now used routinely within the environment and disaster management sectors. These systems allow real world data to be gathered and processed quickly and cost effectively. These spatial video database systems also provide additional value added benefits for an organization since they can also be used for planning or in-house training.  

Data Acquisition

Data acquisition is arguably the most demanding in terms of system complexity and operating environment. The chief objective of data acquisition is to record an image with its corresponding position at a suitable level of positional accuracy and image quality. There is a direct trade off between the two main components namely video and navigation in terms of system cost/complexity and resulting data quality.

At the heart of the system is a video camera, which can be analogue or digital. The camera is usually mounted outside the vehicle, aircraft, boat or locomotive at a suitable position and attitude to maximize the amount of image information captured. Video can be recorded at varying rates of between 1 frame per second (fps) and 25 fps using different image sizes and formats. Video can be composed of analogue, interlace images recorded at 25 fps using S-VHS or Hi8 recorders or full-frame, progressive-scan digital images at rates of 5 fps. GPS and other navigation systems are used to record vehicle position. These sensors can also record at different rates and positional accuracies. GPS is usually available at three levels; standard level with selective availability (SA) switched off, differential (D)GPS or kinematic GPS. Some source of timing is required to synchronise video and navigation data streams. This ensures that an image is tagged with an appropriate navigation message.

Simplified schematic of video-GPS encoding system

Data Processing

Data  processing is required before any useful spatial information can be extracted from video data. This area covers analogue to digital conversion of video, decoding navigation data from video, coordinate projection and final construction of spatial database. A frame-grabbing card is used to digitize video if an analogue camera system was used during the acquisition phase. XY coordinates which, were collected and stored with the video, need to be extracted and converted into the local grid system. Finally, both the video and navigation data streams need to be structured using an industry standard GI file format system such as ESRI shape files or MapInfo tables. Other data processing steps include camera calibration and camera model generation where the relationship between video image and map space is defined.

Data Browsing

Data browsing enables the user to sift through the many tens of thousands of video frames and carry out measurements and build asset databases in a cost effective manner. The browser can operate in ‘stand alone’ mode on a single desktop or ‘client-server’ mode where many people can access the central image data store. The basic data model is shown in the figure below. The browser comprises three integrated modules; video, database and map or GI module.  

Mapping based on Ordnance Survey of Ireland No. NE0001402

©Ordnance Survey Ireland and Government of Ireland