Systematic Classification of Curvature and Feature Descriptor of 3D Shape and Its Application to "Complexity" Quantification Methods

Generative design is a system that automates part of the design process, but it cannot evaluate psychological issues related to shapes, such as "beauty" and "liking". Designers therefore evaluate and choose the generated shapes based on their experience. Among the design features, "complexity" is considered to influence "aesthetic preference". Although feature descriptors calculated from curvature can be used to quantify "complexity", the selection guidelines for curvature and feature descriptors have not been adequately discussed. Therefore, this study aimed to conduct a systematic classification of curvature and a feature descriptor of 3D shapes and to apply the results to the "complexity" quantification. First, we surveyed the literature on curvature and feature descriptors and conducted a systematic classification. To quantify "complexity", we used five curvatures (Gaussian curvature, mean curvature, Casorati curvature, shape index, and curvature index) and a feature descriptor (entropy of occurrence probability) obtained from the classification and compared them with the sensory evaluation values of "complexity". The results showed that the determination coefficient between the quantified and sensory evaluation values of "complexity" was highest when the mean curvature was used. In addition, the Casorati curvature tended to show the highest signal-to-noise ratio (i.e., a high determination coefficient irrespective of the parameters set in the entropy calculation). These results will foster the development of generative design of 3D shapes using psychological evaluation.

Study on the Automatic Identification of ABX3 Perovskite Crystal Structure Based on the Bond-Valence Vector Sum

Perovskite materials have a variety of crystal structures, and the properties of crystalline materials are greatly influenced by geometric information such as the space group, crystal system, and lattice constant. It used to be mostly obtained using calculations based on density functional theory (DFT) and experimental data from X-ray diffraction (XRD) curve fitting. These two techniques cannot be utilized to identify materials on a wide scale in businesses since they require expensive equipment and take a lot of time. Machine learning (ML), which is based on big data statistics and nonlinear modeling, has advanced significantly in recent years and is now capable of swiftly and reliably predicting the structures of materials with known chemical ratios based on a few key material-specific factors. A dataset encompassing 1647 perovskite compounds in seven crystal systems was obtained from the Materials Project database for this study, which used the ABX3 perovskite system as its research object. A descriptor called the bond-valence vector sum (BVVS) is presented to describe the intricate geometry of perovskites in addition to information on the usual chemical composition of the elements. Additionally, a model for the automatic identification of perovskite structures was built through a comparison of various ML techniques. It is possible to identify the space group and crystal system using just a small dataset of 10 feature descriptors. The highest accuracy is 0.955 and 0.974, and the highest correlation coefficient (R2) value of the lattice constant can reach 0.887, making this a quick and efficient method for determining the crystal structure.

Matching Algorithm for 3D Point Cloud Recognition and Registration Based on Multi-Statistics Histogram Descriptors

Establishing an effective local feature descriptor and using an accurate key point matching algorithm are two crucial tasks in recognizing and registering on the 3D point cloud. Because the descriptors need to keep enough descriptive ability against the effect of noise, occlusion, and incomplete regions in the point cloud, a suitable key point matching algorithm can get more precise matched pairs. To obtain an effective descriptor, this paper proposes a Multi-Statistics Histogram Descriptor (MSHD) that combines spatial distribution and geometric attributes features. Furthermore, based on deep learning, we developed a new key point matching algorithm that could identify more corresponding point pairs than the existing methods. Our method is evaluated based on Stanford 3D dataset and four real component point cloud dataset from the train bottom. The experimental results demonstrate the superiority of MSHD because its descriptive ability and robustness to noise and mesh resolution are greater than those of carefully selected baselines (e.g., FPFH, SHOT, RoPS, and SpinImage descriptors). Importantly, it has been confirmed that the error of rotation and translation matrix is much smaller based on our key point matching algorithm, and the precise corresponding point pairs can be captured, resulting in enhanced recognition and registration for three-dimensional surface matching.

Evaluation of 2D-/3D-Feet-Detection Methods for Semi-Autonomous Powered Wheelchair Navigation

Powered wheelchairs have enhanced the mobility and quality of life of people with special needs. The next step in the development of powered wheelchairs is to incorporate sensors and electronic systems for new control applications and capabilities to improve their usability and the safety of their operation, such as obstacle avoidance or autonomous driving. However, autonomous powered wheelchairs require safe navigation in different environments and scenarios, making their development complex. In our research, we propose, instead, to develop contactless control for powered wheelchairs where the position of the caregiver is used as a control reference. Hence, we used a depth camera to recognize the caregiver and measure at the same time their relative distance from the powered wheelchair. In this paper, we compared two different approaches for real-time object recognition using a 3DHOG hand-crafted object descriptor based on a 3D extension of the histogram of oriented gradients (HOG) and a convolutional neural network based on YOLOv4-Tiny. To evaluate both approaches, we constructed Miun-Feet—a custom dataset of images of labeled caregiver’s feet in different scenarios, with backgrounds, objects, and lighting conditions. The experimental results showed that the YOLOv4-Tiny approach outperformed 3DHOG in all the analyzed cases. In addition, the results showed that the recognition accuracy was not improved using the depth channel, enabling the use of a monocular RGB camera only instead of a depth camera and reducing the computational cost and heat dissipation limitations. Hence, the paper proposes an additional method to compute the caregiver’s distance and angle from the Powered Wheelchair (PW) using only the RGB data. This work shows that it is feasible to use the location of the caregiver’s feet as a control signal for the control of a powered wheelchair and that it is possible to use a monocular RGB camera to compute their relative positions.

Comparative Analysis of Detectors and Feature Descriptors for Multispectral Image Matching in Rice Crops

Precision agriculture has greatly benefited from advances in machine vision and image processing techniques. The use of feature descriptors and detectors allows to find distinctive keypoints in an image and the use of this approach for agronomical applications has become a widespread field of study. By combining near infrared (NIR) images, acquired with a modified Nikon D80 camera, and visible spectrum (VIS) images, acquired with a Nikon D300s, a proper crop identification could be obtained. Still, the use of different sensors brings an image matching challenge due to the difference between cameras and the possible distortions from each imaging technique. The aim of this paper is to compare the performance of several feature descriptors and detectors by comparing near infrared and visual spectral bands in rice crop images. Therefore, a group of 20 different scenes with different cameras and growth stages in a rice crop were evaluated. Thus, red, green, blue (RGB) and L, a, b (CIE L*a*b*) channels were extracted from VIS images in order to compare the matches obtained between each of them and the corresponding NIR image. The BRISK, SURF, SIFT, ORB, KAZE, and AKAZE methods were implemented, which act as descriptors and detectors. Additionally, a combination was made between the FAST algorithm for the detection of keypoints with the BRIEF, BRISK, and FREAK methods for features description. BF and FLANN matching methods were used. The algorithms were implemented in Python using OpenCV library. The green channel presented the highest number of correct matches in all methods. In turn, the method that presented the highest performance both in time and in the number of correct matches was the combination of the FAST feature detector and the BRISK descriptor.

Realworld 3D Object Recognition Using a 3D Extension of the HOG Descriptor and a Depth Camera

3D object recognition is an generic task in robotics and autonomous vehicles. In this paper, we propose a 3D object recognition approach using a 3D extension of the histogram-of-gradients object descriptor with data captured with a depth camera. The presented method makes use of synthetic objects for training the object classifier, and classify real objects captured by the depth camera. The preprocessing methods include operations to achieve rotational invariance as well as to maximize the recognition accuracy while reducing the feature dimensionality at the same time. By studying different preprocessing options, we show challenges that need to be addressed when moving from synthetic to real data. The recognition performance was evaluated with a real dataset captured by a depth camera and the results show a maximum recognition accuracy of 81.5%.

Large-Scale Assessment of Bioinformatics Tools for Lysine Succinylation Sites

Lysine succinylation is a form of posttranslational modification of the proteins that play an essential functional role in every aspect of cell metabolism in both prokaryotes and eukaryotes. Aside from experimental identification of succinylation sites, there has been an intense effort geared towards the development of sequence-based prediction through machine learning, due to its promising and essential properties of being highly accurate, robust and cost-effective. In spite of these advantages, there are several problems that are in need of attention in the design and development of succinylation site predictors. Notwithstanding of many studies on the employment of machine learning approaches, few articles have examined this bioinformatics field in a systematic manner. Thus, we review the advancements regarding the current state-of-the-art prediction models, datasets, and online resources and illustrate the challenges and limitations to present a useful guideline for developing powerful succinylation site prediction tools.

Automated Detection of Firearms and Knives in a CCTV Image

Closed circuit television systems (CCTV) are becoming more and more popular and are being deployed in many offices, housing estates and in most public spaces. Monitoring systems have been implemented in many European and American cities. This makes for an enormous load for the CCTV operators, as the number of camera views a single operator can monitor is limited by human factors. In this paper, we focus on the task of automated detection and recognition of dangerous situations for CCTV systems. We propose algorithms that are able to alert the human operator when a firearm or knife is visible in the image. We have focused on limiting the number of false alarms in order to allow for a real-life application of the system. The specificity and sensitivity of the knife detection are significantly better than others published recently. We have also managed to propose a version of a firearm detection algorithm that offers a near-zero rate of false alarms. We have shown that it is possible to create a system that is capable of an early warning in a dangerous situation, which may lead to faster and more effective response times and a reduction in the number of potential victims.

Early prediction of medication refractoriness in children with idiopathic epilepsy based on scalp EEG analysis

Refractory epilepsy often has deleterious effects on an individual's health and quality of life. Early identification of patients whose seizures are refractory to antiepileptic drugs is important in considering the use of alternative treatments. Although idiopathic epilepsy is regarded as having a significantly lower risk factor of developing refractory epilepsy, still a subset of patients with idiopathic epilepsy might be refractory to medical treatment. In this study, we developed an effective method to predict the refractoriness of idiopathic epilepsy. Sixteen EEG segments from 12 well-controlled patients and 14 EEG segments from 11 refractory patients were analyzed at the time of first EEG recordings before antiepileptic drug treatment. Ten crucial EEG feature descriptors were selected for classification. Three of 10 were related to decorrelation time, and four of 10 were related to relative power of delta/gamma. There were significantly higher values in these seven feature descriptors in the well-controlled group as compared to the refractory group. On the contrary, the remaining three feature descriptors related to spectral edge frequency, kurtosis, and energy of wavelet coefficients demonstrated significantly lower values in the well-controlled group as compared to the refractory group. The analyses yielded a weighted precision rate of 94.2%, and a 93.3% recall rate. Therefore, the developed method is a useful tool in identifying the possibility of developing refractory epilepsy in patients with idiopathic epilepsy.

Efficient and robust model-to-image alignment using 3D scale-invariant features

This paper presents feature-based alignment (FBA), a general method for efficient and robust model-to-image alignment. Volumetric images, e.g. CT scans of the human body, are modeled probabilistically as a collage of 3D scale-invariant image features within a normalized reference space. Features are incorporated as a latent random variable and marginalized out in computing a maximum a posteriori alignment solution. The model is learned from features extracted in pre-aligned training images, then fit to features extracted from a new image to identify a globally optimal locally linear alignment solution. Novel techniques are presented for determining local feature orientation and efficiently encoding feature intensity in 3D. Experiments involving difficult magnetic resonance (MR) images of the human brain demonstrate FBA achieves alignment accuracy similar to widely-used registration methods, while requiring a fraction of the memory and computation resources and offering a more robust, globally optimal solution. Experiments on CT human body scans demonstrate FBA as an effective system for automatic human body alignment where other alignment methods break down.