Sulfur-fumigated ginger identification via brightness information and voting mechanism

As a worldwide consumed food, ginger is often sulfur-fumigated. Sulfur-fumigated ginger is harmful to health. However, traditional methods to detect the sulphur-fumigated ginger are expensive and unpractical for general public. In this paper, we present an efficient and convenient identification method based on image processing. Firstly, rapid detection kits are employed to mark three levels of sulfur-fumigated gingers, and the RGB images of the gingers of each sulfur-fumigated level are collected. Secondly, the brightness and texture features are extracted from the images. Three machine learning methods, SVM (Support Vector Machine), BPNN (Back Propagation Neural Network) and RF (Radom Forest) are applied to establish prediction models. Thirdly, the accuracy of each model is calculated and different weights are assigned for different models. Finally, the models with different weights vote the result and the final identification model is established. Experimental results show that the proposed method is robust. When the train set occupies 90%, the prediction accuracy is up to 100%. When the train set only occupies 10%, the accuracy remains high at 80%. Meanwhile, the proposed method is more competitive than other methods in terms of accuracy.