Linear Elements Separation via Vision System Feature and Seed Spreading from Topographic Maps

1. Three-dimensional (3D) eye and head rotations were measured with the use of the magnetic search coil technique in six healthy human subjects as they made large gaze shifts. The aims of this study were 1) to see whether the kinematic rules that constrain eye and head orientations to two degrees of freedom between saccades also hold during movements; 2) to chart the curvature and looping in eye and head trajectories; and 3) to assess whether the timing and paths of eye and head movements are more compatible with a single gaze error command driving both movements, or with two different feedback loops. 2. Static orientations of the eye and head relative to space are known to resemble the distribution that would be generated by a Fick gimbal (a horizontal axis moving on a fixed vertical axis). We show that gaze point trajectories during eye-head gaze shifts fit the Fick gimbal pattern, with horizontal movements following straight "line of latitude" paths and vertical movements curving like lines of longitude. However, horizontal (and to a lesser extent vertical) movements showed direction-dependent looping, with rightward and leftward (and up and down) saccades tracing slightly different paths. Plots of facing direction (the analogue of gaze direction for the head) also showed the latitude/longitude pattern, without looping. In radial saccades, the gaze point initially moved more vertically than the target direction and then curved; head trajectories were straight. 3. The eye and head components of randomly sequenced gaze shifts were not time locked to one another. The head could start moving at any time from slightly before the eye until 200 ms after, and the standard deviation of this interval could be as large as 80 ms. The head continued moving for a long (up to 400 ms) and highly variable time after the gaze error had fallen to zero. For repeated saccades between the same targets, peak eye and head velocities were directly, but very weakly, correlated; fast eye movements could accompany slow head movements and vice versa. Peak head acceleration and deceleration were also very weakly correlated with eye velocity. Further, the head rotated about an essentially fixed axis, with a smooth bell-shaped velocity profile, whereas the axis of eye rotation relative to the head varied throughout the movement and the velocity profiles were more ragged. 4. Plots of 3D eye orientation revealed strong and consistent looping in eye trajectories relative to space.(ABSTRACT TRUNCATED AT 400 WORDS)

Image matting is a technique used to extract a foreground object from an input color image by estimating its opacity mask, which is called an alpha matte. Many previous methods have estimated a poor-quality alpha matte when a foreground object and its background have similar colors. To overcome this problem, we treat an image matting problem as a binary classification problem to classify unknown pixels into foreground and background pixels. In this study, we propose a learning-based matting method based on binary classification. In our method, a binary alpha matte is first obtained by a binary classifier. The binary alpha matte is then iteratively refined to a high-quality alpha matte by repeat use of the classifier. Although we used a support vector machine classifier and the closed-form matting method, an important merit of our method is that it provides a general mechanism to refine an alpha matte by combining various binary classifiers and matting methods. The excellent performance of our method was revealed in several experiments, especially, for input images having similar foreground and background colors as well as objects with small holes.

A digital elevation model and a topographic map of an archaeological site in Syria were created from a field topographic survey with a handheld laser range finder and data post‐processing using a notebook personal computer with Geographical Information System software. The method enabled rapid on‐site topographical mapping even under technical and political restrictions in the country. In an area of about 1 km2, coordinates of random points on the land surface were measured, and a DEM with a 10 m interval grid was generated from the point cloud using Kriging interpolation. The DEM enables various topographic representations to be constructed, such as contour lines, cross sections and slope distribution maps. The DEM‐derived contour lines with a 1 m interval were combined with vector data showing roads and buildings to provide a topographic map. Topographic cross sections created from the DEM generally agree with sections surveyed with a hand level and a measurement tape. The obtained DEM and map are useful for preliminary field research.

Automatic extraction and vectorization of contour lines from color topographic maps is an important precursor to obtaining useful information for many vector based GIS applications. In this work, a novel hybridized algorithm is developed for reconstructing the extracted contour lines from color topographic map. The extraction of contour lines from a topographic map leads to broken contour lines due to inherent characteristics of the map, thus posing a challenging problem of identifying gaps and then filling them. This has been addressed by developing algorithms based on connected components, graph theory, Expectation Maximization (EM) and numerical methods. Our algorithm operates by isolating the segments of those contours which have gaps and achieves in reducing the complexity of the matching of such segments by employing the EM algorithm. We also present a new scheme of filling gaps present in thick contours without the application of thinning algorithms.

Landslides and instability slopes are major risks for human activities which often lead to losing economic resources and damaging properties and structures. The main aims of this study are identifying the underlying effective factors of landslide occurrence in Bijar, Kurdistan Province, and evaluating the regions prone to landslide to prepare the susceptibility map using the logistic regression (LR) and analytical hierarchy process (AHP). At first, using field surveys, questionnaires, geological and topographic maps and reviewing the related studies, ten effective factors including the elevation of sea level, slope inclination, slope aspect, geology, distance from the linear elements (fault, road, and river), precipitation and land use were recognized. Then, they were processed using ARC GIS 10 and ILWIS 33. The dependent variable included 144 of slopes prone to landslide selected across the region as the landslide data (code 1), and also 144 stable landslide slopes were randomly selected as landslide free data (code 0). The results of the evaluation showed that LR model with PCPT index equals to 83.4; −2LL index equals to 229.226; and ROC index equals to 98.5% and landslide susceptibility map based on SCAI index had high verification in the case study. Therefore, 75.489% of the area had very low susceptibility, 10.037% low susceptibility, 3.628% moderate susceptibility, 4.062% high susceptibility and 6.784% very high susceptibility. Based on the preferences of the AHP method, the weighting of selected parameters was logically performed so that the parameters could be arranged according to their priorities. The results of the AHP model showed that 3.4% of the area had very low susceptibility, 30.43% low susceptibility, 46.68% moderate susceptibility, 18.14% high susceptibility, and 1.33% very high susceptibility.

The representation of the Earth’ topography in a 3D digital format is a vital constituent in a wide range of geographic, geomatics, environmental, and engineering activities. However, the accuracy of such representations could define its applicability and validity for a specific utilization. This chapter presents an analysis of four open-source Global Digital Elevation Models (GDEMs) and compares them on two topographic profiles (nearly flat, and hills regions) for mapping and geomatics applications in two countries. The main goal is to determine whether GDEMs-based heights, contour intervals, slopes, and topographic profiles are accurate for all topographic map scales, which is a significant problem in mapping operations. For the purpose of illustrating flat and moderate-topography patterns, two case studies—the Nile delta in Egypt and Makkah city in Saudi Arabia—have been used. The investigated GDEMs include the most-recent released models: ASTER v.3, ACE 2, SRTMGL1 v. 3, and NASADEM_HGT v.1 released in 2019 and 2020 with spatial resolutions of 1 and 3 arc seconds. There are 540 accurate Ground Control Points (GCP) available in the Nile delta and 175 in Makkah. It has been determined based on the available datasets in two study locations that  the accuracy of investigated GDEMs over known checkpoints range from ±2.5 and ±5.1 meters in the Nile delta region, while it varies between ±5.1 and ±8.0 meters in the Makkah area. This suggests that the use of GDEMs in topographic mapping varies greatly between spatial areas that are flat and those that are hilly. Therefore, it is advised against utilizing GDEMs to create topographic maps at scales of 1: 25,000 or greater in flat areas and 1: 50,000 or greater in hilly areas. Additionally, the achieved results demonstrated that throughout cross sections up to 30 kilometers in length, no GDEM-based slope matches the real slopes from known GCP. This leads to the conclusion that GDEMs are not the appropriate heights' source for topographic mapping at medium and large map scales, and could not be utilized for topographic profiling in precise engineering and geomatics applications. Such findings should be vitally taken into account in mapping and engineering applications not just in Egypt and Saudi Arabia but worldwide as well.

The use of regional sinkhole maps, or closed-contour depicted sinkholes from 7.5-minute topographic maps, for deriving sinkhole susceptibility maps may be problematical. In Virginia, the 1:24,000-scale topographic maps for the Valley and Ridge physiographic province karst utilize either 6- or 12-m (20- or 40-ft) contour intervals. Detailed field mapping of sinkholes reveals that the actual number of sinkholes is an order of magnitude greater than the number depicted on a 6-m (20-ft) contour topographic map. The majority of the newly-formed Virginia sinkholes, examined over twenty years, were triggered by human influences to local drainage or by induced groundwater fluctuation. In folded and faulted carbonate-rock terrains, sinkhole distributions are influenced by site-specific lithologic configuration, bedrock deformation, and local hydrology. The processes of sinkhole formation are attuned to water-rock-soil relationships. Sinkhole susceptibility mapping should not be based on sinkhole patterns reflective of previous and existing land-use conditions if it will be used to plan for future development and land-use changes that will modify hydrological conditions. Process driven, karst is an evolving terrain that commonly has been characterized by subsidence, but the most significant and pervasive hazard is the susceptibility of karst groundwater to contamination.

The study of the earth mass movement has long been regarded as one of the most important and interesting aspect of engineering geology and geotechnical engineering, which the designers and planners from the private and public sectors address when implementing the initial stage of urban and rural development projects. This involves highways and infrastructures construction and land use planning among the others. Failure to appreciate the problems relating to mass movements of earth material could lead to damage of man made structures and even the loss of lives. These studies focused on the mass movement in Bundu Tuhan to Kundasang highway area approximately 84 km to 96 km from Kota Kinabalu city, Sabah, one of the most vulnerable to mass movements occurrence in west coast of Sabah. It is bounded by longitude line E 116o 31.592’ to E 116o 36.183’ and latitude line N 06o 00.269’ to N 05o 57.610’. The main objectives of this study are; 1) to map and locate the landslides in the study area; and 2) to study the mechanism and the influence of geological factors causing the mass movement. Geology of the study area and its surrounding is hosted mainly by three sedimentary rock formations: Trusmadi Formation (Palaeocene to Eocene age), Crocker Formation (Late Eocene to Early Miocene age) and Pinousuk Gravel (Upper Pleistocene to Holocene age). These three geologic formations dissected by numerous geological lineaments structural produced by a complex tectonic history of multi phase folding and thrust, normal and reverse faulting. These tectonic setting reduce the physical and mechanical properties of the soil and produced intensive displacement in substrata resulting in intensive high degree of weathering processes. The weathered materials are weak and cause sinking, subsidence and sliding due to high pore pressure subjected by both shallow and deep groundwater. Evaluation 10 boreholes data in study area indicated that the groundwater table in study area is shallow and range 1.9 meter to about 11.3 meters. The groundwater in study area fluctuate drastically even within short period. Sand and gravel layer with variable thickness defined the major shallow aquifers within the top weathered materials while the highly fracture sedimentary rocks defined the major deep aquifers. Most of the aquifer within top unconsolidated weathered material is under unconfined condition. Most of significant aquifers within the sedimentary rocks are sandstones. The sandstones generally fracture and contain coarse sediments, which increase the permeability. Geologic and geotechnic evaluation of the study area indicates that the mass movement take place when slope materials are no longer able to resist the force of gravity. These decrease the shear resistance resulting mass movement, which is due to internal and external factors. Internal factors involve some change in either physical or chemical properties of the rock and soil. External factors involve increase of shear stress on slope, which usually involves a form of disturbance that is induced by man. The triggering mechanism in the study area most likely involves heavy rainfall causing water saturation of the slope material and loss of cohesion along rapture planes. The sheared shale, bedding and fault planes, and opening fractures are all structural weaknesses, which acting as pathways for water seepage, hastening the weakening and eventual mass movement in the study area. Development planning has to consider these hazards in order to counter their effect. An environmental management program should be implemented to prevent these losses. Geological and geotechnical studies will play a vital role in ground stability assessment that critical in public safety.

The comprehensive use of high-resolution remote sensing (HRS) images and deep learning (DL) methods can be used to further accurate urban green space (UGS) mapping. However, in the process of UGS segmentation, most of the current DL methods focus on the improvement of the model structure and ignore the spectral information of HRS images. In this paper, a multiscale attention feature aggregation network (MAFANet) incorporating feature engineering was proposed to achieve segmentation of UGS from HRS images (GaoFen-2, GF-2). By constructing a new decoder block, a bilateral feature extraction module, and a multiscale pooling attention module, MAFANet enhanced the edge feature extraction of UGS and improved segmentation accuracy. By incorporating feature engineering, including false color image and the Normalized Difference Vegetation Index (NDVI), MAFANet further distinguished UGS boundaries. The UGS labeled datasets, i.e., UGS-1 and UGS-2, were built using GF-2. Meanwhile, comparison experiments with other DL methods are conducted on UGS-1 and UGS-2 to test the robustness of the MAFANet network. We found the mean Intersection over Union (MIOU) of the MAFANet network on the UGS-1 and UGS-2 datasets was 72.15% and 74.64%, respectively; outperforming other existing DL methods. In addition, by incorporating false color image in UGS-1, the MIOU of MAFANet was improved from 72.15% to 74.64%; by incorporating vegetation index (NDVI) in UGS-1, the MIOU of MAFANet was improved from 72.15% to 74.09%; and by incorporating false color image and the vegetation index (NDVI) in UGS-1, the MIOU of MAFANet was improved from 72.15% to 74.73%. Our experimental results demonstrated that the proposed MAFANet incorporating feature engineering (false color image and NDVI) outperforms the state-of-the-art (SOTA) methods in UGS segmentation, and the false color image feature is better than the vegetation index (NDVI) for enhancing green space information representation. This study provided a practical solution for UGS segmentation and promoted UGS mapping.

Selectivity to orientations of edges is seen at the earliest stages of visual processing in retinal orientation-selective ganglion cells (OSGCs), which are thought to prefer vertical or horizontal orientation. However, because stationary edges are projected on the hemispherical retina as lines of longitude or latitude, how edge orientation is encoded and decoded by the brain is unknown. Here, by mapping the orientation selectivity (OS) of thousands of OSGCs at known retinal locations in mice, we identify three OSGC types whose preferences match two longitudinal fields and a fourth type matching two latitudinal fields, with the members of each field pair being non-orthogonal. A geometric decoder reveals that two OS sensors yield optimal orientation decoding when approaching the deviation from orthogonality we observe for OSGC field pairs. Retinotopically organized decoding generates type-specific variation in decoding efficiency across the visualfield. OS tuning is greater in the dorsal retina, possibly reflecting an evolutionary adaptation to an environmental gradient of edges.

Determining the shapes of landsurfaces from topographical maps

Estimating the correspondence between the images using optical flow is the key component for image fusion, however, computing optical flow between a pair of facial images including backgrounds is challenging due to large differences in illumination, texture, color and background in the images. To improve optical flow results for image fusion, we propose a novel flow estimation method, wavelet flow, which can handle both the face and background in the input images. The key idea is that instead of computing flow directly between the input image pair, we estimate the image flow by incorporating multi‐scale image transfer and optical flow guided wavelet fusion. Multi‐scale image transfer helps to preserve the background and lighting detail of input, while optical flow guided wavelet fusion produces a series of intermediate images for further fusion quality optimizing. Our approach can significantly improve the performance of the optical flow algorithm and provide more natural fusion results for both faces and backgrounds in the images. We evaluate our method on a variety of datasets to show its high outperformance.

Traditional infrared pseudo color codings based on RGB color model are obtained from gray space; therefore, the color of this kind of pseudo color infrared image has a maximum of 256 types, which cannot fully represents the temperature information of the infrared image, also hard to conform to human visual characteristic. In order to solve these problems, this paper proposes a temperature data pseudo color coding method, which is based on perceptual color space and color difference theory and breaks the limitation of the number of gray scales. Experimental results show that the proposed color coding method has a uniform color transition, and the color of the pseudo color image is much richer.

Analysis and Algebra on Differentiable Manifolds

We propose a new model to assess the effectiveness of camouflage in terms of perceived color difference and gradient magnitude. The “image color similarity index” (ICSI) and gradient magnitude similarity deviation (GMSD) were employed to analyze color and texture differences, respectively, between background and camouflage images. Information entropy theory was used to calculate weights for each metric, yielding an overall camouflage effectiveness metric. During the analysis process, both spatial and color perceptions of the human visual system (HVS) were considered, to mimic real-world observations. Subjective tests were used to compare our proposed method with previous methods, and our results confirmed the validity of assessing camouflage effectiveness based on perceived color difference and gradient magnitude.