Abstract- The procedure for transferring written or printed text from an image file or scanned document into a machine-readable format for data processing applications like editing is known as optical character recognition, or OCR. Although handwriting has been a means of communication for millennia, the advent of contemporary computers has made it simpler. Although most individuals are used to inputting words on a keyboard, mathematical expressions and formulae require additional add-ons to be loaded in the word processor. The procedure might take a long time and be tiresome. Consequently, this study suggests an alternate approach that turns handwritten mathematical formulas into text that can be read by computers. Segmentation is accomplished using horizontal and vertical projection, while character recognition is accomplished by convolutional neural networks, which improve recognition accuracy. The suggested technique was able to extract and recognize each character written in addition to segmenting out handwritten mathematical equations from lined paper. It was then possible to correctly convert the handwritten equation to a digital one.