Fujitsu has developed a new video compression solution that will make it easier to analyze high-definition video footage with AI. According to Fujitsu, video data created with its solution is only a tenth of the size of the data created with other video compression technologies.
To generate those results, Fujitsu observed that AI technologies typically focus on key areas of an image during an analysis. The company’s solution places a similar focus on those key areas, and calculates a minimum image quality that is still sharp enough to be picked up and identified when run through an AI algorithm. Fujitsu then compresses the image down to that minimum level, making the video data as small as possible without affecting the AI’s performance.
Fujitsu believes that its solution will help meet the growing demand for video analytics in a 5G environment. To that end, Fujitsu notes that while 5G technologies will make high-quality video footage easier to capture, it won’t necessarily make it easier to process. The solution is expected to reduce the cost of data transmission, making it ideal for companies that do not need to analyze footage in real time, or are dealing with data from multiple sources.
Fujitsu is hoping to commercialize the technology before the end of 2020, with its COLMINA service platform for the manufacturing industry likely to be one of the first applications. However, the technology will eventually be applied in retail and other fields.
In the meantime, Fujitsu should have plenty of opportunities to take advantage of its own technology. The company recently secured the first private 5G license in Japan. The license covers Fujitsu’s Shin-Kawasaki Technology Square, and the company is already looking to secure licenses for additional locations.
A new report from VIAVI found that 5G networks are now up and running in 378 cities around the world.