Water Quality Image Recognition based on SVM

It is proposed a method that combines digital image processing technology with machine learning algorithm. Firstly, python is used to extract the features of water sample image, the data were pre-processed, divided into training and test sets, and the SVM is used to construct water quality classification models. In the work of classifier evaluation, we need to combine the classifier model performance evaluation indicators: accuracy, Recall, F-value. According to the test results, the accuracy of SVM is more than 90%.