Infrared Image Superpixel Segmentation Based on Seed Strategy of Contour Encoding

Superpixel segmentation has become a crucial pre-processing tool to reduce computation in many computer vision applications. In this paper, a superpixel extraction algorithm based on a seed strategy of contour encoding (SSCE) for infrared images is presented, which can generate superpixels with high boundary adherence and compactness. Specifically, SSCE can solve the problem of superpixels being unable to self-adapt to the image content. First, a contour encoding map is obtained by ray scanning the binary edge map, which ensures that each connected domain belongs to the same homogeneous region. Second, according to the seed sampling strategy, each seed point can be extracted from the contour encoding map. The initial seed set, which is adaptively scattered based on the local structure, is capable of improving the capability of boundary adherence, especially for small regions. Finally, the initial superpixels limited by the image contour are generated by clustering and refined by merging similar adjacent superpixels in the region adjacency graph (RAG) to reduce redundant superpixels. Experimental results on a self-built infrared dataset and the public datasets BSD500 and 3Dircadb demonstrate the generalization ability in grayscale and medical images, and the superiority of the proposed method over several state-of-the-art methods in terms of accuracy and compactness.

Hippocampus-Based Dynamic Functional Connectivity Mapping in the Early Stages of Alzheimer’s Disease

Background: The hippocampus with varying degrees of atrophy was a crucial neuroimaging feature resulting in the declining memory and cognitive function in Alzheimer’s disease (AD). However, the abnormal dynamic functional connectivity (DFC) in both white matter (WM) and gray matter (GM) from the left and right hippocampus remains unclear. Objective: To explore the abnormal DFC within WM and GM from the left and right hippocampus across the different stages of AD. Methods: Current study employed the OASIS-3 dataset including 43 mild cognitive impairment (MCI), 71 pre-mild cognitive impairment (pre-MCI), and matched 87 normal cognitive (NC). Adopting the FMRIB’s Integrated Registration and Segmentation Tool, we obtained the left and right hippocampus mask. Based on above hippocampus mask as seed point, we calculated the DFC between left/right hippocampus and all voxel time series within whole brain. One-way ANOVA analysis was performed to estimate the abnormal DFC among MCI, pre-MCI, and NC groups. Results: We found that MCI and pre-MCI groups showed the common abnormalities of DFC in the Temporal_Mid_L, Cingulum_Mid_L, and Thalamus_L. Specific abnormalities were found in the Cerebelum_9_L and Precuneus of MCI group and Vermis_8 and Caudate_L of pre-MCI group. In addition, we found that DFC within WM regions also showed the common low DFC for the Cerebellum anterior lobe-WM, Corpus callosum, and Frontal lobe-WM in MCI and pre-MCI group. Conclusion: Our findings provided a novel information for discover the pathophysiological mechanisms of AD and indicate WM lesions were also an important cause of cognitive decline in AD.

40-Hz Blue Light Changes Hippocampal Activation and Functional Connectivity Underlying Recognition Memory

Research on light modulation has typically examined the wavelength, intensity, and exposure time of light, and measured rhythm, sleep, and cognitive ability to evaluate the regulatory effects of light variables on physiological and cognitive functions. Although the frequency of light is one of the main dimensions of light, few studies have attempted to manipulate it to test the effect on brain activation and performance. Recently, 40-Hz light stimulation has been proven to significantly alleviate deficits in gamma oscillation of the hippocampus caused by Alzheimer’s disease. Although this oscillation is one of the key functional characteristics of performing memory tasks in healthy people, there is no evidence that 40-Hz blue light exposure can effectively regulate brain activities related to complex cognitive tasks. In the current study, we examined the difference in the effects of 40-Hz light or 0-Hz light exposure on brain activation and functional connectivity during a recognition memory task. Through joint augmentation of visual area activation, 40-Hz light enhanced brain areas mostly in the limbic system that are related to memory, such as the hippocampus and thalamus. Conversely, 0-Hz light enhanced brain areas mostly in the prefrontal cortex. Additionally, functional connection analysis, with the hippocampus as the seed point, showed that 40-Hz light enhanced connection with the superior parietal lobe and reduced the connection with the default network. These results indicate that light at a frequency of 40 Hz can change the activity and functional connection of memory-related core brain areas. They also indicate that in the use of light to regulate cognitive functions, its frequency characteristics merit attention.

An effective approach for CT lung segmentation using region growing

X-ray is an important means of detecting lung diseases. With the increasing incidence of lung diseases, computer-aided diagnosis technology is of great significance in clinical treatment. It has become a hot research direction to use computer-aided diagnosis to recognize chest radiography images, which can alleviate the uneven status of regional medical level. For clinical diagnosis, medical image segmentation can enable users to timely obtain the target region they are interested in and analyze it, which is significant to be used as an important basis for auxiliary research and judgment. In this case, a region growing algorithm based on threshold presegmentation is selected for lung segmentation, which integrates image enhancement, threshold segmentation, seed point selection and morphological post-processing, etc., to improve the segmentation effect, which also has certain reference value for other medical image processing.

Building Extraction from Terrestrial Laser Scanning Data with Density of Projected Points on Polar Grid and Adaptive Threshold

The extraction of building information with terrestrial laser scanning (TLS) has a number of important applications. As the density of projected points (DoPP) of facades is commonly greater than for other types of objects, building points can be extracted based on projection features. However, such methods usually suffer from density variation and parameter setting, as illustrated in previous studies. In this paper, we present a building extraction method for single-scan TLS data, mainly focusing on those problems. To adapt to the large density variation in TLS data, a filter using DoPP is applied on a polar grid, instead of a commonly used rectangular grid, to detect facade points. In DoPP filtering, the threshold to distinguish facades from other objects is generated adaptively for each cell by calculating the point number when placing the lowest building in it. Then, the DoPP filtering result is further refined by an object-oriented decision tree mainly based on grid features, such as compactness and horizontal hollow ratio. Finally, roof points are extracted by region growing on the non-facade points, using the highest point in each facade cell as a seed point. The experiments are conducted on two datasets with more than 1.7 billion points and with point density varying from millimeter to decimeter levels. The completeness and correctness of the first dataset containing more than 50 million points are 91.8% and 99.8%, with a running time of approximately 970 s. The second dataset is Semantic3D, of which the point number, completeness and correctness are about 1.65 billion, 90.2% and 94.5%, with a running time of about 14,464 s. The test shows that the proposed method achieves a better performance than previous grid-based methods and a similar level of accuracy to the point-based classification method and with much higher efficiency.

Differential Privacy Protection of Face Images Based on Region Growing

Face images, as an information carrier, are rich in sensitive information. Direct publication of these images would cause privacy leak, due to their natural weak privacy. Most of the existing privacy protection methods for face images adopt data publication under a non-interactive framework. However, the E-effect under this framework covers the entire image, such that the noise influence is uniform across the image. To solve the problem, this paper proposes region growing publication (RGP), an algorithm for the interactive publication of face images under differential privacy. This innovative algorithm combines the region growing technique with differential privacy technique. The privacy budget E is dynamically allocated, and the Laplace noise is added, according to the similarity between adjacent sub-images. To measure this similarity more effectively, the fusion similarity measurement mechanism (FSMM) was designed, which better adapts to the intrinsic attributes of images. Different from traditional region growing rules, the FSMM fully considers various attributes of images, including brightness, contrast, structure, color, texture, and spatial distribution. To further enhance algorithm feasibility, RGP was extended to atypical region growing publication (ARGP). While RGP limits the region growing direction between adjacent sub-images, ARGP searches for the qualified sub-images across the image, with the aid of the exponential mechanism, thereby expanding the region merging scope of the seed point. The results show that our algorithm can satisfy E-differential privacy, and the denoised image still have a high availability.

The mechanism of digitized landscape architecture design under edge computing

The purposes are to analyze the mechanism of digitized landscape architecture design and stablize the garden landscape image display in constructing garden landscape digitization platform. According to previous research and mobile edge computing, a scheme of digitized landscape architecture design is proposed based on edge computing. This scheme uses discrete elevation calculation to preserve the landscape design image’s frame. It adopts the Roberts edge detection and Laplacian operator for high-level stable preservation of landscape images. Simultaneously, the displayed image is stablized using edge computing algorithms. Simulation experiments are performed to verify the effectiveness of the proposed scheme of digitized landscape architecture design based on mobile edge computing. Results demonstrate that the discrete elevation calculation algorithm can avoid low visual rendering in the 3D image generation process, optimize the seed point matching of edge correlation, and ensure image clarity and stability. The edge computing algorithm can fundamentally avoid the problem of image shaking. The impact of different algorithm models on the classification and accuracy of landscape images is analyzed through parameter optimization. Compared with some latest models, the proposed landscape design scheme based on edge computing has better accuracy. The average accuracy can reach more than 90%, and the Kappa coefficient remains at 86.93%. The designed garden landscape digitization platform can stably display 3D garden landscape images while avoiding the shaking of 3D images, which can provide a theoretical basis and practical value for designing and planning landscape architecture.

Ground Segmentation Algorithm of Lidar Point Cloud Based on Ray-Ransac

Aiming at the problems of poor segmentation effect, low efficiency and poor robustness of the Ransac ground segmentation algorithm, this paper proposes a radar segmentation algorithm based on Ray-Ransac. This algorithm combines the structural characteristics of three-dimensional lidar and uses ray segmentation to generate the original seed point set. The random sampling of Ransac algorithm is limited to the original seed point set, which reduces the probability that Ransac algorithm extracts outliers and reduces the calculation. The Ransac algorithm is used to modify the ground model parameters so that the algorithm can adapt to the undulating roads. The standard deviation of the distance from the point to the plane model is used as the distance threshold, and the allowable error range of the actual point cloud data is considered to effectively eliminate the abnormal points and error points. The algorithm was tested on the simulation platform and the test vehicle. The experimental results show that the lidar point cloud ground segmentation algorithm proposed in this paper takes an average of 5.784 milliseconds per frame, which has fast speed and good precision. It can adapt to uneven road surface and has high robustness.

TUNNEL MODELING USING MOBILE MAPPING LIDAR POINTS

In this paper, a method to model a tunnel using lidar points is presented. The data used was collected using Leica Pegasus Two Ultimate with a Z+F 9012 Profiler mounted on a mobile platform. The tunnel was approximately 151 m long. Visual inspection of a cross-section of the tunnel showed two rail tracks supported on ballast and sidewalks along both sidewalls of the tunnel. The walls and the ceiling of the tunnel were made of five planar surfaces. The tunnel alignment was straight, without any horizontal or vertical curves. The bearing of the central axis of the tunnel was N12.2oW. The following methodology was developed to model just the planar surfaces of the tunnel by excluding the rails, ballast, sidewalks, powerlines, and other accessories. The entire methodology was divided into three broad parts. In the first part, a model cross-section was created. Since the design plans of the tunnel were not available, the model cross-section polyline was created using mean tunnel dimensions from random cross-section points. The model cross-section consisted of the walls and the ceiling of the tunnel. Points were placed at every 1 cm along the model polyline. Six of the model points that represented the shape of the tunnel were selected as salient points. The lower-left salient point was considered as the seed point. In the second part, to define a reference axis of the tunnel, an approximate centerline was manually defined by selecting points at its start and end. Lidar points within 1 m at the start and the end of the tunnel were modeled using the model points to determine the centroids. The reference axis was determined by connecting the centroids at the start and the end of the tunnel. In the third part, the tunnel points were sliced along the reference axis at 5 cm intervals. The model cross-section was matched to points within each tunnel slice using a three-stage approach. In the first stage, the pattern of salient points was matched to the tunnel points by placing the seed point at every tunnel point location. The distances between salient points and their nearest tunnel points were calculated. Ten sets of tunnel points with the least differences to the salient points were shortlisted. In the second stage, a dense point-to-point matching was performed between the model and sliced tunnel data at the shortlisted points. The shortlisted point location with the least difference between the tunnel and the model points was considered as a match. At this point, the model points were hinged to the tunnel points at the seed point location. Hence, in the last stage, a six-parameter affine transformation was performed to match the model points to the tunnel data. The transformed model points at every 5 cm of the length of the tunnel were considered as current shape of the tunnel.

Spine and Individual Vertebrae Segmentation in Computed Tomography Images Using Geometric Flows and Shape Priors

The surgical treatment of injuries to the spine often requires the placement of pedicle screws. To prevent damage to nearby blood vessels and nerves, the individual vertebrae and their surrounding tissue must be precisely localized. To aid surgical planning in this context we present a clinically applicable geometric flow based method to segment the human spinal column from computed tomography (CT) scans. We first apply anisotropic diffusion and flux computation to mitigate the effects of region inhomogeneities and partial volume effects at vertebral boundaries in such data. The first pipeline of our segmentation approach uses a region-based geometric flow, requires only a single manually identified seed point to initiate, and runs efficiently on a multi-core central processing unit (CPU). A shape-prior formulation is employed in a separate second pipeline to segment individual vertebrae, using both region and boundary based terms to augment the initial segmentation. We validate our method on four different clinical databases, each of which has a distinct intensity distribution. Our approach obviates the need for manual segmentation, significantly reduces inter- and intra-observer differences, runs in times compatible with use in a clinical workflow, achieves Dice scores that are comparable to the state of the art, and yields precise vertebral surfaces that are well within the acceptable 2 mm mark for surgical interventions.