Accurately classifying colorectal polyps, or differentiating malignant from benign ones, has a significant clinical impact on early detection and identifying optimal treatment of colorectal cancer. Convolution neural network (CNN) has shown great potential in recognizing different objects (e.g. human faces) from multiple slice (or color) images, a task similar to the polyp differentiation, given a large learning database. This study explores the potential of CNN learning from multiple slice (or feature) images to differentiate malignant from benign polyps from a relatively small database with pathological ground truth, including 32 malignant and 31 benign polyps represented by volumetric computed tomographic (CT) images. The feature image in this investigation is the gray-level co-occurrence matrix (GLCM). For each volumetric polyp, there are 13 GLCMs, computed from each of the 13 directions through the polyp volume. For comparison purpose, the CNN learning is also applied to the multi-slice CT images of the volumetric polyps. The comparison study is further extended to include Random Forest (RF) classification of the Haralick texture features (derived from the GLCMs). From the relatively small database, this study achieved scores of 0.91/0.93 (two-fold/leave-one-out evaluations) AUC (area under curve of the receiver operating characteristics) by using the CNN on the GLCMs, while the RF reached 0.84/0.86 AUC on the Haralick features and the CNN rendered 0.79/0.80 AUC on the multiple-slice CT images. The presented CNN learning from the GLCMs can relieve the challenge associated with relatively small database, improve the classification performance over the CNN on the raw CT images and the RF on the Haralick features, and have the potential to perform the clinical task of differentiating malignant from benign polyps with pathological ground truth.

Shape or color based moment invariants are conventional pattern sensitive features in the object recognition and image description. However, the existing moment invariants are not robust because they handle simplified cases, such as single impact of shape transformation, single color transformation, and a simple combination of them. In this paper, we propose a kind of shape-color moment invariants (SCMIs), taking complicated shape and color transformation into account. It is applicable to more complicated cases, such as changing viewpoints, different illuminations, different camera white balance modes, and different color spaces. The integral-based SCMI construction framework is extensible and general. We evaluate the technique through comprehensive experiments.

This paper proposes an extended 3-D shape retrieval feature, which is based on a new scheme. In our method, continuous principal component analysis is used to align 3-D models into a canonical position. We then represent 3-D polygonal mesh as a 3-D multi-layer closed curve. It can capture the more feature of the 3-D models. The feature of the 3-D model is extracted from the 3-D closed curve by the Fourier transform. And the experiments show it is more efficient.

In this paper, two evaluation factors are proposed to evaluate subdivision methods based on surface moments. Both factors are derived from moment invariants, and mainly used to measure shape stabilization. The experiment results show these factors are able to measure the distortion characteristic of 3-D models.

3D face data have been increased recently with the development of 3D data acquisition technology. 3D face has extensive applications. A proper coordinate system defined on faces is sometimes necessary. However, building a canonical coordinate system for all the 3D faces presenting the physiological vertical direction of human face is still an open question. In this paper, we propose a new 3D face canonical coordinate system based on the physiological structure of human face. Once locating the nose tip and finding the symmetry plane, we use the circular points on the face around the cheekbones to decide the front view, as the direction of the average normal of them is almost vertical to the center axis of human head. The experiment shows that our method is robust to rough data and provides a uniform description for all the 3D face models with different positions and expressions.

We convert watertight polygonal models to volumetric models by using voxelization methods. Then we use invariant feature descriptors such as volume moment invariants and shape histogram for shape retrieval. The experiment result reveals that 3D models represented by voxels are highly robust to noise and perturbations.

3D face scans often contain too much noise and other irregularities in them. Geometric processing techniques such as smoothing and hole filling are employed to preprocess this kind of data. Each face model is then segmented to get a series of partial surfaces and curves. We use surface moment invariants and curve moment invariants as shape descriptors for the shape retrieval of them.