Developer Stories: PhotoMath from MicroBlink

    Many world publications have written about this application. And not just technical ones like Engadget , Techcrunch and CNET . A small app came to the attention of such giants as Time magazine , Forbes and Business Insider and the Ellen DeGeneres show.



    So why is PhotoMath so interesting? Perhaps because the Croatian MicroBlink development teamI realized our childhood dream by handing a magic wand to solve even complex mathematical equations (in fact, it’s not at all harmful, but, on the contrary, useful for teaching mathematical sciences, but more on that later). Our colleagues managed to contact developers and talk about how this small but successful application was created, and what we can expect from them in the future. In the meantime, your child is burning the last volume of Ready Homework, let me invite you to read excerpts from this interview.

    But first, let's talk a little more about the application:


    The free PhotoMath application was conceived as a useful and visual tool for solving mathematical equations, which is based on the technology of recognition of data (OCR-technology) coming from the camera of a mobile device.



    All you need to do to calculate the equation is to point the smartphone’s camera at the task, select the boundaries of the equation (you can manually adjust it), take a picture of the example and wait until the solution appears on the screen. App can scan tasks both from paper and from the screen, but in the second case it may take more time to process.

    We make a reservation at once that the handwritten text is not yet recognized and many users are indignant at the reviews in the application store in anticipation of the addition of this function from the developers.



    The educational part of the application is that the application gives not only the correct answer, but also allows you to view all the stages of solving a mathematical equation. You can return to all examples resolved with the application at any time, since they are automatically stored in a special section.



    You can try PhotoMath for Lumia smartphones in the above example, after downloading it from the Windows Phone Store . The application also exists in the version for iOS , and will soon appear on the Android platform.

    What inspired you to develop PhotoMath?

    We developed mobile OCR technology and applications for several clients from the banking sector, designed to transfer payments and pay bills. The technology of text recognition implemented in these programs allowed users, using the smartphone’s camera, to transfer data to the device and make payment in a matter of seconds. At some point, we wanted to expand the application of our development to other areas, but we did not have a real demo application.

    One of our colleagues had the idea that [since he often has to check homework for his son] he needs an application for quickly solving mathematical problems, and that it would be nice to create such an application. No sooner had this idea dispersed throughout our team, as we had already applied for a local event held as part of AppCampus . Surprisingly, literally on the first attempt, we won their hearts and after a few months received a grant for the development of the application.



    With the growing popularity of your application, more and more talk has begun to appear that PhotoMath will allow you to "cheat the system" or even completely turn off the head of schoolchildren and students, doing all the work for them. What do you think about it?

    We rightly believe that anyone who wants to cheat will always find a way to cheat using any device or cheat sheet. Using the PhotoMath application, we want to help children and their parents learn mathematics in a new way - using an application that parses the solution of a specific example in detail.

    Agree, it often happens that students have no one to ask for help in solving problems in mathematics, or the knowledge of parents in mathematics is not enough. It is in such cases that our PhotoMath application should come to the rescue.

    What did the development process look like?

    The development process took place conditionally in two stages. At the first, we developed universal OCR technology. And it took us 2-3 years. And then about six months later it took to develop a mathematical engine and the application itself.



    Three young developers from our team worked directly on the application, and we, three “old-timers” only monitored the creation of the product and controlled the development process.

    What application would you develop if you had unlimited resources and budget?

    Since all our company’s attention is focused on the technology of image processing on portable devices, we have many ideas in this area. The first application that we would like to implement was a mobile app that could recognize food, or rather, various dishes. Such an application could offer users a variety of options, from preparing content for publication on social networks and ending with the output of up-to-date statistics on calories and the content of certain substances in a dish.


    Here is another example of our fantasy flight: a robot that autonomously moves around your home or office and remembers all the things around you. What is it for? Yes, so that users could ask the robot or mobile application connected to this robot: “Where did I get my glasses?”

    And such an application could save lives: today people often walk along the street, staring at their mobile phones. Crossing the road, they have great chances not only to become the culprits of the accident, but also to suffer in it. Imagine that there would be such an application (a kind of mobile guardian angel) that would use your mobile camera and microphone to inspect the space around you and notify you of any obstacles on the way. For example, if you have a step in front of you or you are approaching an intersection.

    What awaits in the near future PhotoMath and your other applications?

    Today, we are fully focused on improving PhotoMath and our OCR technology. As for the mathematical application, in the near future we will work on the automatic calculation of more complex calculations and the improvement of the recognition system (maybe someday we can handle the handwriting). If we talk about the next steps of our company, now we are working on the release of the "machine vision" SDK so that other developers can create their applications using our recognition technology.



    You can download the PhotoMath application to your Lumia here .


    And we write on: Habrahabr | Twitter | VK | Facebook | Google+

    Also popular now: