Remember meForgot password?
    Log in with Twitter

article imageWhen analyzing images big data beats the brain

By Tim Sandle     Mar 27, 2017 in Technology
When it comes to viewing and interpreting a series of images it seems that big data computing approaches are superior to the human brain. New findings have implications for the way images, such as those from CCTV, are interpreted.
An important application of the research is with improving traffic safety, having better health screening, and profiling the environmental. Each of these areas can be advanced with image analysis and recognition technology.
Developing advanced image technology requires computer systems to mimic the human brain. Our brains are very good at distinguishing significant from unimportant information and computer programs find it difficult to replicate this activity. This has been the main stumbling block in developing big data approaches, according to Eirik Thorsnes at Uni Research, which is located in Bergen, Norway.
Dr. Thorsnes has spent many years working on a project to develop advanced computing power designed to work in a similar way to the human brain, but to do so faster and with greater accuracy. In a statement, Dr. Thorsnes explains: "There has been a tremendous development in recent years, and we are now surpassing the human level in terms of image recognition and analysis.”
This fits in with the ‘big data’ paradigm. Big data is a term for large data sets, so complex that traditional data processing application software is inadequate for processing the accumulated information. Challenges to dealing with big data include data capture, storage, analysis; together with data curation, search, sharing, transfer, visualization, querying, updating and information privacy.
Dr. Thorsnes has developed algorithms that can notice even the tiniest nuances that humans cannot see. The developed software can analyze large volumes of images and video in a relatively short time. The system also uses artificial intelligence to that learn to recognize objects and recognize those objects that are important in the image. This is a slow process, and the research group has said training a computer is similar to training a child to learn how to read.
The aim is to apply the image analysis system to such areas as environmental monitoring (such as monitoring slopes susceptible to landslides), seabed surveys (including applying imaging to assess the health of fish farms), and satellite images. Another area is medicine, using computers to look for signs of illness. With traffic analysis reviewing the types of vehicles that pass through a particular area at any given point in time can help with traffic planning, especially deciding the routes that business traffic should take.
More about big data, Brain, Scanning, Images, CCTV
More news from
Latest News
Top News