Accurate shade image matching by using a smartphone camera.

PURPOSE: Dental shade matching by using digital images may be feasible when suitable color features are properly manipulated. Separating the color features into feature spaces facilitates favorable matching. We propose using support vector machines (SVM), which are outstanding classifiers, in shade classification. METHODS: A total of 1300 shade tab images were captured using a smartphone camera with auto-mode settings and no flash. The images were shot at angled distances of 14-20cm from a shade guide at a clinic equipped with light tubes that produced a 4000K color temperature. The Group 1 samples comprised 1040 tab images, for which the shade guide was randomly positioned in the clinic, and the Group 2 samples comprised 260 tab images, for which the shade guide had a fixed position in the clinic. Rectangular content was cropped manually on each shade tab image and further divided into 10×2 blocks. The color features extracted from the blocks were described using a feature vector. The feature vectors in each group underwent SVM training and classification by using the "leave-one-out" strategy. RESULTS: The top one and three accuracies of Group 1 were 0.86 and 0.98, respectively, and those of Group 2 were 0.97 and 1.00, respectively. CONCLUSIONS: This study provides a feasible technique for dental shade classification that uses the camera of a mobile device. The findings reveal that the proposed SVM classification might outperform the shade-matching results of previous studies that have performed similarity measurements of ΔE levels or used an S, a*, b* feature set.

Improving point correspondence in cephalograms by using a two-stage rectified point transform.

BACKGROUND: An improved point correspondence method was developed for automatically detecting two-dimensional cephalometric landmarks. The proposed method uses a two-stage rectified point transform: the global correspondence of interest points between two images and the local correspondence of landmarks. METHOD: In the first stage, point-to-point matching pairs were established using local corner point features. The matched points on an input image were treated as a set of transformations, with varying directions and magnitudes, from the template image. Similarity of the transformation vectors was achieved through rectification to exclude vectors that deviated widely from the statistical mean. Rectification attempted to remove noise and irrelevant matched points. In the second stage, the point correspondences were fine-tuned within the regional centers of the landmarks, which were classified into three categories-corners, edges, and structural points-and each category was fine-tuned using a different strategy. Correspondence was performed by evaluating the shortest Euclidean distance between the point descriptors of the template and test images. RESULTS: The correspondence results of 20 orthodontic landmarks were compared with those identified by dental professionals on 80 digital cephalograms collected from a dental clinic. The proposed method detected both hard and soft tissue landmarks with mean error distances of 1.63mm, compared with the 2-mm standard reported by previous studies. CONCLUSIONS: This study enhanced the point correspondence technique for cephalometric landmarking. Using the proposed method, users can preferentially and flexibly add and remove landmarks on a template before correspondence without intensive image pretraining.