Mingjian Hong

Robust image corner detection based on scale evolution difference of planar curves

In this paper, a new corner detector is proposed based on evolution difference of scale pace, which can well reflect the change of the domination feature between the evolved curves. In Gaussian scale space we use Difference of Gaussian (DoG) to represent these scale evolution differences of planar curves and the response function of the corners is defined as the norm of DoG characterizing the scale evolution differences. The proposed DoG detector not only employs both the low scale and the high one for detecting the candidate corners but also assures the lowest computational complexity among the existing boundary-based detectors. Finally, based on ACU and Error Index criteria the comprehensive performance evaluation of the proposed detector is performed and the results demonstrate that the present detector allows very strong response for corner position and possesses a better detection and localization performance and robustness against noise.

Compressed sensing MRI using Singular Value Decomposition based sparsity basis

Magnetic Resonance Imaging (MRI) is an essential medical imaging tool limited by the data acquisition speed. Compressed Sensing is a newly proposed technique applied in MRI for fast imaging with the prior knowledge that the signals are sparse in a special mathematic basis (called the ‘sparsity’ basis). During the exploitation of the sparsity in MR images, there are two kinds of ‘sparsifying’ transforms: predefined transforms and data adaptive transforms. Conventionally, predefined transforms, such as the discrete cosine transform and discrete wavelet transform, have been adopted in compressed sensing MRI. Because of their independence from the object images, the conventional transforms can only provide ideal sparse representations for limited types of MR images. To overcome this limitation, this work proposed Singular Value Decomposition as a data-adaptive sparsity basis for compressed sensing MRI that can potentially sparsify a broader range of MRI images. The proposed method was evaluated by a comparison with other commonly used predefined sparsifying transformations. The comparison shows that the proposed method could give a sparser representation for a broader range of MR images and could improve the image quality, thus providing a simple and effective alternative solution for the application of compressed sensing in MRI.

Joint Local Regressors Learning for Face Alignment

Cascaded regression approaches have been widely applied to computer vision tasks recently, and achieve state-of-the-art performance. In this paper, we consider the problem of face alignment model fitting using cascaded regression and propose a new face alignment approach named Joint Local Regressors Learning (JLRL). The main novelty of our learning framework lies in two following aspects: (1) shape constraints among facial landmarks are considered by jointly learning local regressors; (2) the contribution to face alignment errors for each facial landmark is explored in the training phase. Compared with the previous face alignment methods that have shown state-of-the-art performances, our JLRL approach performed best on the LFPW, Helen and 300-W datasets which are the most challenging datasets today.

Weighted Fusion of Multiple Models for Wavelength Selection

A new method based on the weighted fusion of multiple models is presented for wavelength selection in multivariate calibration of spectral data. It fuses the regression coefficients of multiple models with weights based on minimum mean square error to improve the accuracy and stability of the wavelength selection. To validate the performance of the proposed method, it was applied to the partial least squares (PLS) modeling of three near-infrared spectral datasets and compared with full-spectrum PLS, genetic algorithm–based PLS, and uninformative variable elimination–based PLS methods. Results show that the proposed method can effectively select the informative wavelength and enhance the prediction ability of the PLS model. On account of its simpler algorithm and higher efficiency, it can be widely used in practical applications.

Online dynamic cardiac imaging based on the elastic-net model

Abstract Purpose: The purpose of this work was to develop an online dynamic cardiac MRI model to reconstruct image frames from partial acquisition of the Cartesian k-space data, which utilizes structural knowledge of consecutive image frames. Materials and methods: Using an elastic-net model, the proposed algorithm reconstructs dynamic images using both L1 and L2 norm operations. The L1 norm enforces the sparsity of the frame difference, while the L2 norm with motion-adaptive weights catches the internal structure of frame differences. Unlike other online methods such as the Kalman filter (KF) technique, the new model requires no assumption of Gaussian noise, and can faithfully reconstruct the dynamic images within a compressive sensing framework. Results: The proposed method was evaluated using simulated dynamic phantoms with 40 frames of images (128 × 128) and a cardiac MRI cine of 25 frames (256 × 256). Both results showed that the new model offered a better performance than the online KF method in depicting simulated phantom and cardiac dynamics. Conclusion: It is concluded that the proposed imaging model can be used to capture a large variety of objects in motion from highly under-sampled k-space data, and being particularly useful for improving temporal resolution of cardiac MRI.

Distributed Compressed Sensing MRI Using Volume Array Coil

The volume array coil in the magnetic resonance imaging (MRI) system is a typical application of the distributed sensor network in the biomedical area. Each coil provides a large coverage of the imaged object, and the signals are largely overlapped during the data acquisition. The intercoil image similarities can be explored for the distributed compressed sensing (CS) based image reconstruction. In this work, a singular value decomposition (SVD) based sparsity basis was developed for the CS-MRI with a volume array coil configuration. In this novel imaging method, the spatial correlation both of intracoil and intercoil exploited. The experimental results showed that is with eightfold undersampled k-space data acquisition, the target images could still be faithfully reconstructed using the proposed method, which offered a better imaging performance compared to conventional CS schemes.

Corner detection using multi-scale representation of contour orientation rate based on B-spline: Corner detection using multi-scale representation of contour orientation rate based on B-spline

Multi-scale representation for contour orientation rate was defined based on B-spline scale space,and a new corner detection algorithm was proposed by exploring multi-scale contour orientation rate and its multiplication.The local extrema of multi-scale multiplication were directly defined as candidate corners.Multi-scale multiplication not only included the feature information of orientation rates at several scales,but also could enhance the contour orientation rates of corners while suppressing the rates of contour points corrupted by boundary or quantization noise.Thus,the good corner detection results could be obtained only through using a global threshold and that multi-scale detection of the candidate corners was truly performed.Finally,quantitative analysis of corner detection localization performance was performed,and the comparing experiments illustrate that the proposed algorithm possesses good detection and localization performance.