Soil analysis using machine learning
Agriculture is the backbone of Pakistan’s economy and millions of people are directly or indirectly linked with the agriculture sector. The main source for agriculture is soil for crops production. If soil is healthy then crops production automatically boosts up. But in Pakistan average farmers cannot afford to test their soils in the laboratory which is quite time-consuming and costs a lot. So in this work, we try to facilitate the farmers with an application that can check the quality of soil by using their mobile phone images and recommend different fertilizers according to the soil sample results. About 1064 images are used for machine learning models training with laboratory-tested labels. Different machine learning models like ANN, CNN, Decision Tree Regression, Random Forest Regression, and Support Vector Regression models are trained to predict different nutrients of the soil. After machine learning model training, one of the best models will be deployed on the mobile application to facilitate the farmers' use of machine learning to predict soil nutrients.
Keywords: Machine Learning,Deep Learning,Android Application Development
Tools: Scikit-learn,Tensorflow,Keras,Pandas,Python,Numpy,OpenCV,Java
Department: Department of Computer Science
Project Team Members
Name |
Email |
Mahmood Yousaf
|
yousaf2018@namal.edu.pk |
Project Poster