Point-cloud coordinate information derived from terrestrial Light Detection And Ranging (LiDAR) is important for several applications in surveying and civil engineering. Plane fitting and segmentation of target-surfaces is an important step in several applications such as in the monitoring of structures. Reliable parametric modeling and segmentation relies on the underlying quality of the point-cloud. Therefore, understanding how point-cloud errors affect fitting of planes and segmentation is important. Point-cloud intensity, which accompanies the point-cloud data, often goes hand-in-hand with point-cloud noise. This study uses industrial particle boards painted with eight different colors (black, white, grey, red, green, blue, brown, and yellow) and two different sheens (flat and semi-gloss) to explore how noise and plane residuals vary with scanning geometry (i.e., distance and incidence angle) and target-color. Results show that darker colors, such as black and brown, can produce point clouds that are several times noisier than bright targets, such as white. In addition, semi-gloss targets manage to reduce noise in dark targets by about 2–3 times. The study of plane residuals with scanning geometry reveals that, in many of the cases tested, residuals decrease with increasing incidence angles, which can assist in understanding the distribution of plane residuals in a dataset. Finally, a scheme is developed to derive survey guidelines based on the data collected in this experiment. Three examples demonstrate that users should consider instrument specification, required precision of plane residuals, required point-spacing, target-color, and target-sheen, when selecting scanning locations. Outcomes of this study can aid users to select appropriate instrumentation and improve planning of terrestrial LiDAR data-acquisition.

Laser scanning intensity captures information about the reflectance of target-surfaces and is used for a variety of applications such as data registration, classification, object detection and recognition. To enhance its utility, intensity values often undergo a correction process, which reduces the influence of nuisance parameters on recorded intensity (often range and incidence angle). This study applies and tests the Torrance-Sparrow model to correct intensity for the incidence angle effect in terrestrial laser scanning. Main components of the Torrance-Sparrow model are the geometrical attenuation (G) and microfacet distribution functions (D). Four models of geometrical attenuation and microfacet distribution functions are evaluated, namely, (i) Beckmann, (ii) Trowbridge-Reitz, (iii) GGX, and (iv) shifted gamma distribution (SGD). These models provide different derivations of the functions G and D, which estimate parameters necessary for the Torrance-Sparrow model. Target-surfaces scanned from various incidence angles are used for the assessment. These are painted with eight different colours (white, yellow, red, green, blue, grey, brown, and black) and two sheens (flat and semi-gloss), which create different reflection characteristics (diffuse and specular). Numerical and visual evaluations show that all four models manage to model the specular reflection component of the semi-gloss sheen target-surfaces for all tested colours. However, in flat-sheen surfaces, the Beckmann and SGD models show inferior modelling than GGX and Trowbridge-Reitz for brown, grey, and black colours. In addition, a relative comparison of the roughness parameters and Fresnel factors showed that only the Trowbridge-Reitz model produced reasonable values, based on encountered surface characteristics. Application of the Trowbridge-Reitz model in independent point-cloud data shows how intensity values can be corrected for the incidence angle effect in real cases.

Technological advancements in virtual reality have opened the door for widespread dissemination and application in engineering education. In surveying engineering, outdoor labs are essential, because they offer great means for experiential learning and preparation for real applications. However, challenges related to weather, inaccessibility to sites, transportation costs, and liability reduce time spent with instruments, and limit comprehension of surveying procedures and preparedness for the job market. Through virtual reality, we can create realistic surveying scenarios that are difficult to implement in practice, such as surveying in cities, at construction sites, and in different terrain scenarios. In addition, virtual reality can be used to support remote and online learning. This paper presents the thorough implementation of virtual reality labs in a surveying engineering first-year class. Technical feedback showed that surveying labs were recreated with a high level of fidelity, which is important to transfer skills learned in virtual reality to the real environment. Students stated that virtual reality gave them more incentives to learn, added to the fun of learning, and improved their overall learning experience. Furthermore, students indicated that virtual reality labs helped them understand surveying procedures and the operation of instruments, and they considered virtual labs to be a useful tool for preparation for physical labs. Pedagogical assessments using grades from previous years showed that the virtual reality labs enhanced student learning. Barriers of virtual reality, such as nausea and motion sickness, limit their extensive use. Despite a few negatives, virtual reality offers viable solutions to challenges in surveying engineering education, with the potential for further development and integration in other surveying courses.

In recent years, immersive virtual reality has been used in disciplines such as engineering, sciences, and education. Point-cloud technologies such as laser scanning and unmanned aerial systems have become important for creating virtual environments. This paper discusses creating virtual environments from 3D point-cloud data suitable for immersive and interactive virtual reality. Both laser scanning and sUAS point-clouds are utilised. These point-clouds are merged using a custom-made algorithm that identifies data gaps in the master dataset (laser scanner) and fills them with data from a slave dataset (sUAS) resulting in a more complete dataset that is used for terrain modelling and 3D modelling of objects. The terrain and 3D objects are then textured with custom-made and free textures to provide a sense of realism in the objects. The created virtual environment is a digital copy of a part of the Penn State Wilkes-Barre campus. This virtual environment will be used in immersive and interactive surveying laboratories to assess the role of immersive virtual reality in surveying engineering education.

Applications of convolutional neural network (CNN)-based object detectors in agriculture have been a popular research topic in recent years. However, complicated agricultural environments bring many difficulties for ground truth annotation as well as potential uncertainties for image data quality. Using YOLOv4 as a representation of state-of-the-art object detectors, this study quantified YOLOv4’s sensitivity against artificial image distortions including white noise, motion blur, hue shift, saturation change, and intensity change, and examined the importance of various training dataset attributes based on model classification accuracies, including dataset size, label quality, negative sample presence, image sequence, and image distortion levels. The YOLOv4 model trained and validated on the original datasets failed at 31.91% white noise, 22.05-pixel motion blur, 77.38° hue clockwise shift, 64.81° hue counterclockwise shift, 89.98% saturation decrease, 895.35% saturation increase, 79.80% intensity decrease, and 162.71% intensity increase with 30% mean average precisions (mAPs) for four apple flower bud growth stages. The performance of YOLOv4 decreased with both declining training dataset size and training image label quality. Negative samples and training image sequence did not make a substantial difference in model performance. Incorporating distorted images during training improved the classification accuracies of YOLOv4 models on noisy test datasets by 13 to 390%. In the context of apple flower bud growth-stage classification, except for motion blur, YOLOv4 is sufficiently robust for potential image distortions by white noise, hue shift, saturation change, and intensity change in real life. Training image label quality and training instance number are more important factors than training dataset size. Exposing models to test-image-alike training images is crucial for optimal model classification accuracies. The study enhances understanding of implementing object detectors in agricultural research.