Detection of impurity content in wheat based on cone-beam CT

Wheat impurity content is an essential index for evaluating the quality level and price of wheat or reserved grain. We obtain 2D slice images of the scanned sample by cone-beam CT, then a hybrid filtering algorithm is used to improve the quality of the 2D slice images. Finally, the 2D slice image is reconstructed in 3D to obtain volume data, so as to obtain the volume of each part, and then use the density formula to achieve the calculation of wheat impurity content. The experiments have shown that the proposed method can obtain the detection accuracy up to 90%, and can provide objective online detection of grain impurity content compared with the impurity detection of the national standard GT/T5494-2019.