Crowoptic: Big numbers know what people like to shoot

From a quiet consumer application to the enterprise large data application market, Crowoptic gives us the revelation mainly has two: first, the enterprise-level market is today the hot spot of venture capital; Secondly, in the trend of it consumption and large data, a consumer-oriented start-up company can also flexibly turn to the enterprise-level market.

CrowdOptic, a Silicon Valley start-up company, which has been set up just two years ago, announced that its large data-picture analysis technology has been able to analyze the highest-profile shots in a massive picture of a major public-recorded event.

With the popularity of personal digital photography equipment and the rapid development of technology, the Internet photo video and other rich media data is growing exponentially every year, along with Hurricane Sandy or the Olympic Games such a major event, social networking sites surging users upload tens of millions of photos, these photos contain metadata data, Then it becomes the analysis object of crowdoptic image large data technology.

During the London Olympics, for example, Crowoptic analyzed 1000 of photos uploaded by tourists on social networking sites and found that the ArcelorMittal Mittal Orbit Sightseeing tower outside the Olympic Park was the most photographed target.

The method of crowdoptic is to analyze the metadata in the EXIF information of pictures, including GPS position, compass direction and time stamp. Then, the triangulation technique and algorithm are used to compute an important value for each photo object, which takes only 1.531 seconds. In the same way, crowdoptic the thousands of photos of Hurricane Sandy on social networking sites, analyzing the places and objects most often photographed by amateur photographers.

Shift from consumer to enterprise

CrowdOptic initially hoped to develop a mobile phone app to analyze the lens of the camera, the same object can also initiate real-time conversation between the camera. For example, when a large number of cell phone lenses are targeted at the same location or player in a football match, CrowdOptic's focus technology can give notice to the photographer, allowing them to initiate discussions on a common subject. But after more than a year of product operation, CrowdOptic found that app is too difficult to promote, consumers to shoot and share photos of the habit is very difficult to change. So Crowdoptics decided to do a alternative, through the analysis of social networking site images of metadata data to carry out the popularity of the object analysis. Today, many social networking sites have pictures of crowdoptics needed for triangulation analysis (metadata) (but it's worth noting that the images on Twitter contain a much smaller proportion of metadata). )

To get rid of its reliance on mobile apps, CrowdOptic's technology now captures 1 billion of images from the Internet, analyzing photos containing metadata data and finding the most popular images. CrowdOptic not only can be based on GPS location information, but also on the photograph of the object of large data analysis. Fisher, founder of CrowdOptic, said the technology could be used to classify and correlate images of websites that generate a huge amount of images on a daily basis. Fisher is also bullish on CrowdOptic's application in digital eyewear technology such as Google Glass, where people can see the focus of attention in their surroundings.

Enterprise Search is the direction

Although CrowdOptic's technology sounds cool, the prospects for commercialization are not yet clear. Fisher had previously created a security technology company based on situational analysis, which was sold to Oracle in 2007. In recent years, as the focus of corporate acquisitions from consumer companies to corporate technology companies, crowdoptic future direction is likely to become a large enterprise data analysis and search solutions provider, as Fisher said, today's enterprises are being flooded with pictures and videos, The retrieval of unstructured data is becoming a huge challenge, and that is why HP buys autonomy,oracle why it buys Endeca. But Fisher did not disclose much more about how crowdoptic from the corporate-level market nuggets.

(Responsible editor: The good of the Legacy)