Quality assessment towards cell diffraction image based on multi-channel feature fusion

Image quality assessment towards cell diffraction image is significant for both the academic and medical domain. It plays an important role in medical detection and recognition, such as cell morphology and heterogeneity classification. However, cell diffraction image quality assessment is still a challenging task due to the high heterogeneity of cells and various appearance. To solve this problem, we propose a cell diffraction image quality assessment. More specifically, we first collect cell diffraction images including Jurkat and Ramos. To remove cell impurity and debris images, we leverage the K-means clustering algorithm and support vector machine (SVM) to eliminate these images. Subsequently, we calculate the Gray Level Co-occurrence Matrix (GLCM) of each image and extract deep representation by using DNN. Afterward, we fuse luminance, contrast, GLCM, and deep representation to calculate the feature similarity between the reference image and the test image. Extensive experiments conducted on Jurkat cells and Ramos cells datasets have shown the effectiveness of our proposed method.