I have been given a PC project at High school that is supposed to calculate equations from an image. Obviously i will need some form of image recognition to do this, so i searched the internet for quite some time. I have found two ways to solve this. First one including an external program like Neuroph(a program working with neural networks) and the other being a Java OCR project found here .
The problem i have with Neuroph is that i would have to split the equation image into multiple other images, each containing one character, since the program compares the images in whole. I found this a little tricky to do so i wanted to see if there was a different solution.
And so i came upon the OCR project. Unfortunately i couldn't find much information of its possibilities on the internet, so i am unsure if it can solve the problem i had with Neuroph (having to split the image)
In conclusion i have the following questions:
- Can the Java OCR detect characters on the image?(the splitting problem)
- Does anyone know of a good way to split an image into multiple others, each holding one character? (I can do the rest with Neuroph)
- Is there a 3rd way to make the project?
You can use this SDK to do it easily (I know it works on Android, iOS, and Windows Phone but I'm not sure about other OSes). You will need to contact the company for the maths SDK though. Also, if you want to see how it works, search for 'PhotoMath' on your phone's store.