Emerging of mobilephones has changed the world around us, and those things that 20 years ago seemed fantastic are real today. Nowadays, almost every person has a mobile camera in his pocket that takes high-resolution and high-quality images.
Also changed the pattern of behavior of people and business. Employees have become more mobile, many of them work at home or in public spaces. The scanner is not always at hand in order to make a high-quality image of the document and send it to the business application, so you have to take pictures of documents, such as invoices, using mobile cameras. But this can be a problem if you need to make an OCR of this mobile image. Images obtained from a mobile camera often have flaws that are not typical for images obtained from scanners. Such flaws can significantly degrade the quality of OCR and lead to data loss. Next, we will discuss the main problems that can occur with OCR of images obtained from a mobile camera.
Text lines are bent along with the surface of the page when using a mobile camera. This can be a significant difficulty for the OCR engine, reducing the quality of recognition.
It’s almost impossible to get a perfectly rectangular page when shooting with a mobile camera. Such distortions can also significantly degrade the quality of OCR.
Mobile cameras have different resolutions and this can be a problem, because OCR engines produce maximum quality when all images have a constant resolution, for example, 300 dpi.
Unwanted background and foreign objects in the frame
Often, when using a mobile camera, in addition to the document itself, the surrounding space gets into the frame. Sometimes such objects can be misunderstood by the OCR engine, which may lead to a decrease in the quality of recognition. Such photos need to be cropped to avoid potential issues.
Photos obtained from a mobile camera can be made in such a way that the entire text will be skewed at a certain angle. If you try to make an OCR of such a photo, the results are likely to be bad. Such photos should be deskewed before OCR processing.
There are also other problems encountering when you use images obtained from a mobile camera for OCR. In the next article, we will discuss solutions for this whole complex of problems, and how ABBYY FlexiCapture 12 can help with this.
Ilya Evdokimov is a long-term practitioner and expert in leading Optical Character Recognition (OCR), Data Capture and Document Processing techniques, technologies and solutions. With over 15 years of experience spanning enterprise software implementations, mobile applications development, cloud-based systems integration and desktop-level automation, Ilya Evdokimov uses through industry knowledge and experience to achieve high efficiency and workflow optimization in most challenging paper-dependent and digital image capture environments.