Realizing the Effective Detection of Tumor in Magnetic Resonance Imaging using Cluster-Sparse Assisted Super-Resolution

Recently, significant research has been done in Super-Resolution (SR) methods for augmenting the spatial resolution of the Magnetic Resonance (MR) images, which aids the physician in improved disease diagnoses. Single SR methods have drawbacks; they fail to capture self-similarity in non-local patches and are not robust to noise. To exploit the non-local self-similarity and intrinsic sparsity in MR images, this paper proposes the use of Cluster-Sparse Assisted Super-Resolution. This SR method effectively captures similarity in non-locally positioned patches by training on clusters of patches using a self-adaptive dictionary. This method of training also leads to better edge and texture detection. Experiments show that using Cluster-Sparse Assisted Super-Resolution for brain MR images results in enhanced detection of lesions leading to better diagnosis.

Texture Detection of Aluminum Foil Based on Top-Hat Transformation and Connected Region Segmentation

A new method of texture detection for aluminum foil based on digital image processing technology is proposed. Top-hat transformation and image segmentation technology based on the connected domain are used to change the method of determining texture fraction by using human experience. Compared with the brightness method, pit detection method, and EBSD technology, this method can complete quantitative detection efficiently, automatically, and accurately, and reduce the detection time and manpower. It eliminates the instability of manual detection and ensures the accuracy of detection. By this method, the error of test results can be controlled within 1.6%, which is much better than 7.3% of the brightness method and 4% of the pitting method. It provides more accurate test results for the production process control of aluminum foil.

3D-conductive pathway written on leather for highly sensitive and durable electronic whisker

A leather-based e-whisker with 3D conductive pathway was assembled by writing conductive ink on leather. The sensor possessed good durability, and was sensitive enough to detect a height difference of 50 μm, making it capable of surface texture detection, spatial distribution mapping, wind mapping, etc.

Limitations of CNNs for Approximating the Ideal Observer Despite Quantity of Training Data or Depth of Network

The performance of a convolutional neural network (CNN) on an image texture detection task as a function of linear image processing and the number of training images is investigated. Performance is quantified by the area under (AUC) the receiver operating characteristic (ROC) curve. The Ideal Observer (IO) maximizes AUC but depends on high-dimensional image likelihoods. In many cases, the CNN performance can approximate the IO performance. This work demonstrates counterexamples where a full-rank linear transform degrades the CNN performance below the IO in the limit of large quantities of training data and network layers. A subsequent linear transform changes the images’ correlation structure, improves the AUC, and again demonstrates the CNN dependence on linear processing. Compression strictly decreases or maintains the IO detection performance while compression can increase the CNN performance especially for small quantities of training data. Results indicate an optimal compression ratio for the CNN based on task difficulty, compression method, and number of training images.