To demonstrate how modern technologies can percept real objects, we developed a special MeterSnap project based on Machine Learning. It is an Android application which recognizes meters, and captures readings and serial numbers via a mobile phone camera. MeterSnap captures the readings and a serial number, and then tries to find a pre-registered user in the database to add a record about fixation of the readings to the user’s history. Using this application, it’s possible to make a fixation of subscriber’s readings just in a few simple steps. All you need is a mobile phone with a camera which can distinguish details in poor lighting.
The process of perception is very complex and not fully understandable even for humans. So, try to imagine how difficult it is to learn a machine which hasn’t a brain. Our team of developers accepted a great challenge!
It was necessary to learn the networks to detect a recognition object and serial number on the image, and select the character area. Then, the app should confirm that each meter belongs to a particular person. The client also wanted to introduce a feature of app profiling to work with different meters.
We used the method of a contour analysis that narrows the number of meters, which the application can potentially work with, to presence of a key contour ("anchor"). It is an object that has a unique contour (in most cases, a black box) around the numbers with the readings.
Based on the geometry of the anchor, we implemented a technology which calculates the location of the serial number and accurately recognizes it.
Registering the location of a particular meter in the database, tracking the shooting location, and comparing all known meter data, the app can easily determine the owner of a certain meter.
We transferred all the parameters, which the recognition process may depend from, to the configuration file, and determined parameters for each supported meter with the possibility to choose an operational meter. This procedure allowed us to introduce the useful app profiling feature.
Eventually, we overcome all the challenges thanks to our perseverance and ingenuity! :)
As you see, we developed a very useful Android application based on Machine Learning which can easily recognize meters, captures readings and serial numbers. Our customer provided a society with the ability to receive important data on various meters quickly and precisely. Thanks to our effective cooperation, the truth about unlimited cognitive possibilities of machines, particularly about their excellent perception, was discovered!