MatMath
Math is the cradle of all creations, without which the world cannot move an inch. Mathematical equations have become bread and butter for scientist and engineers. Mathematical functions are used in software engineering, civil and mechanical engineering and in many other fields as well. A problem fac
2025-06-28 16:34:05 - Adil Khan
MatMath
Project Area of Specialization Software EngineeringProject SummaryMath is the cradle of all creations, without which the world cannot move an inch. Mathematical equations have become bread and butter for scientist and engineers. Mathematical functions are used in software engineering, civil and mechanical engineering and in many other fields as well. A problem faced by the students is to derive the solution for a matrix problem using pen and paper-based approach that is a time taking process and for a solver to find an exact formula to solve specific matrix is problematic. Our major goal is to bridge the gap between traditional pen-based methods and technologies by developing an android application which solves matrices problems using image processing. There is no such software and application available in the market which uses image processing to solve matrix problems. Matrix related operations like determinants, transpose, ad-joint, inverse, Cramer rule will be solved in step by step manner for user better understanding. While calculating determinants if matrices include any trigonometric terms as matrix elements, then it will be solved by corresponding trigonometric formulas. There are two options for getting image which can either be taken by camera or it can be select from gallery. Selection of image leads to cropping or resizing of an image. Our application supports matrices up to 10*10 dimensions. Matrix recognition is categorized into two classes i.e. computer printed matrix and hand written matrix. Our main focus is on computer printed form. After getting the image from the camera, OCR (optical character recognition) will get the text from the image and our algorithm will solve the required given problem. Our algorithm is written in python language which is integrated to android. User can select whether he/she wants to check the solution in step by step manner. We are also providing some additional features. These features include math dictionary, history and add to favorite. Our application will open a new dimension in solving matrices problems using image processing.
Math is one of the hardest subjects in school, which is why owning a graphing calculator seems like a necessity for students. But what if you could use your smartphone to solve equations by pointing the camera at the problem in your textbook instead of using a graphing calculator? That is the idea behind MatMath. MatMath is a free mobile app that can read and solve mathematical matrices problem using your smartphone camera in real time. MatMath may sound like it is simply helping kids cheat, but the app also provides a step-by-step guide showing how each of the problems are solved. The step-by-step solution is beneficial to students that do not have access to a tutor and struggle with solving matrix problems. Parents can also use the MatMath app to jog their memory when teaching math to their kids. MatMath currently supports basic Cramer rule, Transpose, Adjoint, Inverse and several other methods to find determinant. These are the most common operations of matrix that MatMath can solve. MatMath can handle matrix up to 10*10 dimension and for Cramer rule up to 3*3 dimension. The MatMath app uses optical character recognition (OCR) technology to read the equation and calculates the answers within seconds. There is a frame in the MatMath app that you have to use to insert the captured image of equation. After getting image from camera or gallery the text in the image will extracted and placed in an editable textbox. In case of any mistake, we can also take another image or correct the text from textbox. MatMath’s ability to read math problems has its limitations. MatMath can only read printed text so your handwritten notes cannot be solved by the app. The main objective of this app is to solve your problems related to matrices through built in camera and give solution step-by-step.
Project Implementation MethodEnvision system is designed and implemented in python and android. To extract the text from captured image we are using Google OCR. We are working on creating a model. Model will be created using TensorFlow, to help reduce dimensionality of data with the help of neural network. This model is used to reduce the dimensionality of data as learned from the data can also be used for reducing noise in data. For this project, we are using different dataset to train the model. In this project, we will create two neural network model –one will be trained to perform classification of these digits and another model is used to denoise input data. This model also works to recognize different matrices, for example some people used square bracket around the matrix and some are used simple bracket. From captured image, model recognized the start braces of matrix and read the data until it recognized close braces. Image from the camera can include more than 1 question, for that we have used a Crop Image API, which helps the user to crop the image and get the question. If the question has a bug or question has a small mistake, for that we have included an edit text function, where a user can edit the extracted text. We have developed our algorithms in python language. This application works offline, and python android integration is also done and worked offline. After the editing of the text, the given question will be solved by calling an algorithm from python which will solve the problem in step by step manner. There is a math dictionary added, we have created our own dataset for the math dictionary. User can also give a review or rate the application in play store. We have also included history and add to favorite features which can help user to find more frequently solved questions.
Today's machine has the capability of looking and understanding things, just like human. This is achieved through image processing and machine learning. The primary aim is to allow the computer learn automatically without human intervention. This system is aimed at using this capability for looking new age problem. Mathematical equations have become bread and better for engineers and scientist. Matrices play important role in all branches of science, engineering, social science and management. Now a day's computer program plays an important role in solving equations. There are many applications and websites are available in the market, which can solve a given equation and return result. These website and application are manually, it is a very time-consuming process, and user enter value one by one. If a matrix is 6x6 dimension, then a user enters 36 values. Just think about the scenario where they want to solve the same question. This system is unique in this way and this is a new dimension of using image processing to solve matrix problem and provide solution more than 2/3 methods of single problem if it's is possible.
Technical Details of Final Deliverable- OCR API
- CROP API
- Model Creation
- Algorithms
- Android python Integration
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 55384 | |||
| RAM (8GB) | Equipment | 1 | 6500 | 6500 |
| Publishing on Google playstore | Miscellaneous | 1 | 3857 | 3857 |
| Printing | Miscellaneous | 10 | 150 | 1500 |
| Overheads | Miscellaneous | 1 | 3627 | 3627 |
| Graphics Card(2GB) | Equipment | 1 | 8900 | 8900 |
| Stationary | Equipment | 2 | 500 | 1000 |
| Collection of dataset to train the system | Equipment | 1 | 30000 | 30000 |