GoPro Camera Research Articles

Image enhancement comparison to improve underwater cultural heritage survey

This work aims at presenting an underwater image application to obtain an improved 3D model of cultural assets. In 2016, more than 500 images were acquired by a GoPro Camera with a low resolution of 72 dpi and focal length of 3 mm, without flash and are now used to reconstruct the 3D model of some amphoras of a Roman shipwreck found in Albenga (Italy). We have applied state-of-art image enhancement techniques, such as ACE, CLAHE, LAB and SP algorithms, to improve the quality of underwater images affected by low contrast, poor visibility conditions, not uniform lighting, colour variations, noise and blur effect. The visual quality has been evaluated through quantitative metrics, like average luminance, information entropy, average gradient of image, UCIQE and UIQM. Then, our efforts have been devoted to the dense 3D point cloud generation using a SfM-MVS software. In particular, the 3D reconstruction results are in line with the metric evaluation: in fact, the more accurate 3D objects are obtained from that enhanced dataset with the highest measured image quality.

Extracting nearshore wave properties from video: A new method for coastal estuaries

Wave measurements, such as height and period, are important in understanding estuarine and coastal processes as well as air-sea interactions, which are important in the validation of hydrodynamic models and improvements in the parameterizations needed for atmospheric models. This paper presents and evaluates a relatively inexpensive and portable video-based method of measuring the nearshore wave properties of a coastal estuary. The proposed system uses a GoPro camera along with a set of scripts to perform camera calibration and remove lens distortions. Significant wave heights and peak frequencies are then extracted from the video without the need for the ground control points (objects of known physical location in relation to the camera) as used in previous studies. By not relying on ground control points, the system becomes portable and easy to deploy, allowing the collection of wave data from various locations on the estuary. The estimated significant wave heights and peak frequencies were validated with a sonic wave gauge located in the camera's field of view. Percent error and RMSE were calculated and indicate the proposed method is robust and the errors (estimated at less than 6%) are consistent with other video-based methods that use multiple cameras and ground control points.

Point-of-view technology to teach surgery.

Varied and limited opportunities to assist during operations, hamper surgical procedural training. Furthermore, the trainee's view of the procedure differs from that of the operating surgeon. New solutions could emerge when instruction under direct supervision can be combined with technological advances in surgical practice. This study assesses the feasibility of using the GoPro HERO™ to record operations for educational purposes, from the perspective of the surgical and technical team. A "point-of-view" wearable GoPro camera was used to record standard operations at Tygerberg Academic Hospital, affiliated with Stellenbosch University, in Cape Town. Semi-structured interviews evaluated the practical acceptability of the GoPro in theatre by clinical personnel. The technical team reflected on picture quality, video editing and formatting. Techsmith Camtasia Studio 9® programme was employed for post- recording editing. Eight operations were performed by three surgeons. No interference with operative sterility, procedural length, nor functioning in the operating room was encountered by clinical personnel. Technically, the quality and picture field were adequate. All surgeons found the camera head band tight and the camera heavy. The limited battery life during recording was problematic throughout the study. Wearable cameras allow recordings of surgical procedures from the surgeon's view, for adjunctive educational purposes. This study confirms the feasibility and utility of GoPro to record real time operations without compromising patient care. The videos have cross-discipline potential for teaching within medicine. Creation of a video library as a resource to evaluate the educational value for trainees is envisioned.

3D DATA GENERATION USING LOW-COST CROSS-VIEW IMAGES

Abstract. 3D data generation often requires expensive data collection such as aerial photogrammetric or LiDAR flight. In cases such data are unavailable, for example, areas of interest inaccessible from aerial platforms, alternative sources to be considered can be quite heterogeneous and come in the form of different accuracy, resolution and views, which challenge the standard data processing workflows. Assuming only overview satellite and ground-level go-pro images are available, which we call cross-view data due to the significant view differences, this paper introduces a framework from our project, consisting of a few novel algorithms that convert such challenging dataset to 3D textured mesh models containing both top and façade features. The necessary methods include 3D point cloud generation from satellite overview images and ground-level images, geo-registration and meshing. We firstly introduce the problems and discuss the potential challenges and introduce our proposed methods to address these challenges. Finally, we practice our proposed framework on a dataset consisting of twelve satellite images and 150k video frames acquired through a vehicle-mounted Go-pro camera and demonstrate the reconstruction results. We have also compared our results with results generated from an intuitive processing pipeline that involves typical geo-registration and meshing methods.

Approaching of GOPRO camera embedded on UAV to calculate NDVI for corn’s crop

Nitrogen (N) is a macronutrient used in large quantities in modern agriculture, due to a quantity required by plants. In the corn crop, it mainly acts on grain filling and affects protein content. One of the causes of the inefficiency of the use of nitrogen fertilizers in the corn crop is the difficulty in estimating in a timely manner the need of nitrogen fertilization through soil and / or plant analyzes. Looking for feasibility of the recommendation process of nitrogen fertilization in real time, researches has been developed aiming to detect nitrogen deficiency using remote sensing. In this paper the objective was to evaluate the feasibility of the use of the Normalized Difference Vegetation Index (NDVI), calculated by means of multispectral image processing obtained with the GOPRO’s camera and the Micasense’s Seqouia camera, embedded in UAV’s, to recommend the application of nitrogen fertilizers to the corn crop. According to results obtained with this paper, it was possible to conclude that the determination of NDVI by means of aerial images is compatible with the determination of the index using active optical sensor (GreenSeeker).

Traffic light and moving object detection for a guide‐dog robot

Guide dogs are helpful for visually impaired people for navigating through the streets. However, it is expensive and time consuming to train a guide dog. In addition, a guide dog cannot decide when and where to cross a street safely, and it is up to the human to decide. Here, the authors propose a framework for creating a guide dog robot by using artificial intelligence and other technologies. The proposed framework is based on an Intel UP squared board, together with a Neural Compute Stick Movidius to process the images gathered from a GoPro camera. MobileNet single shot detector (SSD) is the main framework to detect the moving objects in the environment. The final decision is made after fusing the information gathered from all the sources. The authors also apply the Amazon Alexa device for the voice communication between the guide dog robot and the visually impaired person. A prototype of the proposed system is implemented and tested. Experimental results show that the proposed framework can process the information at a traffic intersection scene and can guide a blind person to cross the street safely.

Spatial video remote sensing for urban vegetation mapping using vegetation indices

Urban vegetation is important because of the fast-growing urbanization. If we want cities to have sustainable growth and well-kept ecology, we need to develop a smart and efficient urban vegetation monitoring system. This paper examines the possibility of using a modified GoPro camera mounted on a car. The lens of a GoPro camera was replaced with the NDVI-7 lens to obtain blue, green and near-infrared band. The performance of four vegetation indices was tested: Blue normalized difference vegetation index (BNDVI), Green normalized difference vegetation index (GNDVI), Green-blue normalized difference vegetation index (GBNDVI), Blue-wide dynamic range vegetation index (BWDRVI). Based on those indices, binary classification was performed to classify objects in the scene as either vegetation or non-vegetation. Finally, the accuracy of each index was assessed on three different study sites. Results show that GBNDVI performs best for the given task with average classification accuracy of 95.10% for all study sites.

Unique Hydration Caves and Recommended Photogrammetric Methods for Their Documentation

Active anhydrite hydration during weathering leading to crystallisation of secondary gypsum and significant volume expansion was investigated and documented by the authors at three sites: the environs of Walkenried (Germany), Dingwall (Canada), and Pisky (Ukraine). As a result of these processes, peculiar landscape forms were created: hydration domes and ridges with empty internal chambers, some of them large enough to be called hydration caves (German: Quellungshöhlen). Currently, there are only four recognised sites on Earth featuring such a unique landscape and with a large group of hydration caves in one place (the fourth site is in the Alebastrovyye Islands, Russia). These sites constitute a particularly valuable geological and geomorphological heritage, including potential geosites and geomorphosites which require special protection. Actively growing hydration domes and caves change shape and size within a short time span, on the scale of months, years, or decades. Their study and proper protection require these changes to be monitored. Several different methods of documentation were applied in the field in order to document continuing morphological changes. The practical aspects of the use of each of these methods were assessed, demonstrating that the photogrammetric methods offer the greatest utility; not only are they the most efficient (fast and sufficiently precise) but also, compared with other methods, they yielded the most complete results. The key documentation of outcrops in Canada and Ukraine was executed with the application of terrestrial photogrammetry at Pisky (GoPro camera) and aerial photogrammetry at Dingwall (unmanned aerial vehicle). Application of these methods enabled the recording of the morphology associated with the hydration process in the form of 2.5D and 3D models as well as of orthophotomaps. The maps and the models were created using the Photoscan programme. The authors demonstrate that the photogrammetric models can be used for spatial morphological analysis of hydration forms in the ArcGIS programme. Repetition of this documentation in future will enable analysis of the morphological changes expected to occur during the progressive expansive hydration of anhydrite.

Mat leikskólabarna á þátttöku í tilviksrannsókn

Greinin fjallar um mat og þátttöku leikskólabarna í rannsókn um sköpunarsmiðjur í leikskólanum þeirra og var tilgangurinn að rýna í hvernig börn upplifðu þátttöku í rannsókninni. Bakgrunnur rannsóknarinnar byggir á ákvæði í Barnasáttmála Sameinuðu þjóðanna um þátttöku barna. Gagna var aflað í mars (smiðjur) og júní (rýniviðtal) 2018. Níu fimm ára börn voru þátttakendur í sex vinnusmiðjum þar sem leikið var með stafræn tæki, legokubba og annan skapandi efnivið. Meðan á smiðjum stóð söfnuðu rannsakendur gögnum með fjölbreyttum hætti, svo sem með myndbandsupptökuvél og myndavél og héldu dagbækur, börnin öfluðu gagna með smámyndavél (GoPro), spjaldtölvu, fylltu út matsblað eftir hverja smiðju og rýniviðtal var tekið við þau eftir að smiðjum lauk. Í greininni eru matsblöð barnanna, dagbækur rannsakenda og rýniviðtal við börnin lögð til grundvallar niðurstöðum.Helstu niðurstöður sýna að börn eru verðugir þátttakendur í rannsókn og að sjónarhorn þeirra varpar nýju ljósi á gögnin í rannsókninni. Fram kom að börnin virtust vera ánægðust með þau verkefni sem þau höfðu sjálf vald yfir, hvort sem það sneri að sköpun eða tækjabúnaði, þannig heillaði smámyndavélin sem þau stýrðu sjálf, litlu stafrænu tækin og að taka viðtöl hvert við annað með spjaldtölvu. Hins vegar er vert að benda á að einn af þeim þáttum sem kom fram í niðurstöðum er að sum barnanna upplifðu viðveru rannsakenda í smiðjum stundum truflandi. Það er vert fyrir rannsakendur sem vinna rannsóknir með börnum að hafa hugfast.

Toward Understanding of the Peanut Maturity Classification Process

HighlightsThe manual mesocarp color-based classification process for runner peanuts was investigated to inform the development of an automatic method for peanut maturity classification.The results show that the maturity classification process is not solely dependent on mesocarp color. Other factors, such as the shape of the sample distribution, are considered during the manual process.Color detection and distribution shape should be considered in the future development of automatic peanut maturity classification system.Abstract. Optimizing the harvest date is essential for maximizing peanut yields and ensuring top-quality peanuts. Currently, the most popular and acceptable method for peanut maturity classification is a manual method called the peanut profile board, which is based on the color of the peanut’s mesocarp layer. An automated system could greatly improve the efficiency of the current manual method. However, little research has been done on the manual method, so there is not enough insight to guide the development of an automatic system. In this study, the aim was to analyze the process of using the peanut profile board and gather relevant information for developing an automatic method for peanut maturity classification. Specifically, a GoPro camera was used to record on-site sample classification by trained human classifiers. Classification times were analyzed through the recorded videos, and the classification results were compared with a solely color-based classification method to see if trained classifiers consider factors other than color. With an average classification time of 11.98 s per peanut, the Black 1 color category required more classification time than the other color categories, which indicates that Black 1 is most important for maturity classification. The results from the solely color-based classification method had an average color matching rate of 69.3% with the results of the trained classifiers, which indicates that peanut maturity evaluation is not based only on color. The maturity prediction, defined as the ratio of the number of peanuts in the combined brown and black categories to the total sample size, differed by an average of only 13.4% between the two classification methods. This indicates that mesocarp color is the major factor for peanut maturity classification. Overall, the results of this study prove the feasibility of automated color detection and classification for peanut maturity evaluation. Future development of such a system will benefit from the insights found in this study. Keywords: Color classification, Peanut classification time, Peanut maturity classification procedure.

Evaluation of the Quality of Manual, Compression-Only Cardiopulmonary Resuscitation in a Moving Ski Patrol Toboggan.

Introduction: Cardiopulmonary resuscitation (CPR) quality may be impacted by location and setting of an out of hospital cardiac arrest. This study compared the quality of CPR performed on a moving ski patrol toboggan versus stationary CPR, both performed outdoors in winter. Materials and Methods: Compression-only CPR was performed on a manikin attached to a backboard secured into a toboggan. A CPR device was used to measure compression rate, depth and recoil, and elapsed time. A convenience sample of 30 patrollers, in weather-appropriate clothing, participated in this nonrandomized, crossover study. Each first performed 5 continuous sets of 30 compressions of stationary CPR straddling the manikin while kneeling. After 15 minutes rest, participants performed CPR while moving down the designated ski run (fixed length, vertical drop, and slope angles). Each ski run was video captured with a GoPro camera. Quality was defined as compliance with 2015 International Liaison Committee on Resuscitation (ILCOR) guidelines for CPR. Results: Overall, stationary and moving chest compressions complied with ILCOR guidelines, but there was a statistically significant degradation of CPR quality while moving and over time. Fewer compressions met ILCOR guidelines in comparison to stationary CPR compressions: (1) stationary mean depth 87% compliant, moving mean depth 35% compliant, 95% confidence interval: 39-65 (p < 0.001); (2) stationary mean rate 90% while moving mean rate 64% compliant; and (3) stationary recoil 74%, while moving recoil 77% compliant. Noncompliant compressions were typically too shallow, and noncompliance for rate was typically too fast. There were no pauses over 10 seconds once compressions were started. Conclusions: Despite CPR quality being reduced while moving, there were sufficient compliant compressions to support the use of CPR in this setting. Maintaining regular CPR training in a working environment and optimal body position in relation to the patient may be keys to performing high quality CPR on a moving toboggan.

140 Use of GoPro to Assess Central Venous Line Placement Competency Asynchronously

Complications due to failed central venous access (CVA) attempts are associated with increased morbidity and health care costs. The Agency for Healthcare Research and Quality (AHRQ) estimates a central line associated blood stream infection (CLABSI) costs approximately $70,000. There is a high failure rate in completing ultrasound guided central venous access (UGCVA) procedures among first year resident interns. Training and assessment is important, however physician time to assess resident competency is limited. Previous research has demonstrated that a GoPro camera can be used to capture procedures. In this research, we will evaluate the use of video recordings to asynchronously evaluate resident procedural competency in ultrasound guided central venous line (USGCVL) placement. This research had several steps. First, we evaluated the ability of GoPro cameras to capture procedural competencies during USGCVL procedures. Next, we developed and validated an evaluation rubric consisting of critical competencies identified from the literature and institutional guidelines. Faculty reviewers evaluated the GoPro videos asynchronously for visual clarity, and completeness of observation while using the rubric. Results demonstrated high agreement between reviewers (80-100%) for visualization of each competency. Residents were recruited to participate in this study during a training course designed to assess procedural competency in central line placement using a cadaver. Following consent, participants donned a head-mounted GoPro camera to record their performance. During the procedure, faculty evaluated residents' performance synchronously (via Qualtrics) using the same evaluation rubric validated in the prior study. At a later date, two teams each composed of a surgeon and emergency medicine (EM) physician (different than the synchronous reviewers) asynchronously viewed and independently evaluated videos divided into groups of four using the same rubric via Qualtrics. Inter-rater reliability was calculated between pairs of asynchronous reviewers and average percent agreement was calculated for both the asynchronous and the asynchronous versus live reviews. Video recordings of twenty-eight USGCVL procedures were collected. Twenty-four met study inclusion criteria (1080p 30fps continuous video no significant glare, no gross misalignment of the video field of view) and twenty are included in the data analysis. (Four videos have not been reviewed by their assigned reviewers yet.) Results are presented in Table 1. Overall percent agreement approached 80% for both asynchronous and asynchronous versus live comparisons. This study supports the feasibility of using a head mounted GoPro camera to record video for asynchronous evaluation of resident USGCVL placement. Continued refinement of data capture, scoring practices, and the grading rubric is needed to increase both the validity and reliability of the assessment process.Table 1Average percent agreement.Average Percent Agreement (%)95% CI (%)Asynchronous Rater Comparison 12 VideosRater 2 (surgeon) vs. Rater 3 (EM)73.4744.04100Live Graders vs. Asynchronous 8 VideosRater 1, 3 (EM) vs. live graders78.1242.54100 Open table in a new tab

When friends recommend: online purchasing behavior of Russian and Dutch people when prompted by recommendations from Facebook friends

ABSTRACTThis study investigates the relationship between product recommendations in a Facebook advertisement and the behavioral intentions of consumers, and to what extend this relationship is influenced by cultural differences in the tendency to avoid uncertain situations. In an online experiment 142 Dutch and 92 Russian people (average age 24 years) were presented with two advertisements of a GoPro camera in a Facebook context. The product recommendations in the advertisements differed in tie strength, that is, the amount and type of recommendations by close Facebook friends (strong ties) and other Facebook friends (weak ties). The results showed that the Dutch group had higher purchase intentions than the Russian group. Both groups are more influenced by strong tie online friends, than by weak tie online friends. The tie strength effect was stronger for the Dutch group. Meanwhile, compared to the Dutch group, the Russian group had a higher level of uncertainty avoidance, and also reported that they valued less the recommendations of other people, than those from their online friends. In addition, they were more willing to give the GoPro camera as a present to their close Facebook friends, than to other Facebook friends.

Going along with older people: exploring age-friendly neighbourhood design through their lens

Neighbourhoods are extremely important to older people, as this is where a great deal of their everyday life is spent and where social interaction happens. This is particularly the case in deprived neighbourhoods, where people with limited economic resources or physical limitations find it challenging to venture outside the neighbourhood. A growing body of research suggests studying age-friendly neighbourhoods from a bottom-up approach which takes the diversity of the age group into account. This paper aims to investigate how the go-along method can serve to co-construct knowledge about age-friendly neighbourhood design in a deprived neighbourhood of Copenhagen with a diverse group of older people. Sixteen go-along interviews were carried out with older people aged 59–90. The participants took on an expert role in their own everyday life and guided the researcher through the physical and social environments of their neighbourhood. The go-alongs were documented with a GoPro camera. The data was analysed using situational analysis and was grouped into thematic categories. Our findings conclude that social interaction is the overall motivator for going outdoors and that dimensions of pavements, the seating hierarchy, the purpose of lawns, sheltered spaces and ‘unauthorised’ places are all neighbourhood design elements that matter in this regard. The findings suggest to consider age-friendly details as the starting point for social interaction, to target the appropriate kind of age-friendly programs and to enhance empowerment through physical spaces. The go-along interview as a research method holds the potential for empowering older people and appreciating their diversity.

THE AT (THE AMAZING TECHNOLOGY) MEMPERKENALKAN WILAYAH MARITIM DAN POTENSI DAERAH PROVINSI KEPULAUAN RIAU MENGGUNAKAN VISUAL VIDEO DAN BERBASIS ANDROID SEBAGAI MEDIA PENGUATAN PENGETAHUAN ANAK SEKOLAH DASAR DI PESISIR

Kepulauan Riau (Kepri) is the 32nd province in the Republic of Indonesia which is determined by the DPRRI based on the Constitution No. 25 with Tanjungpinang as the capital city. According to the Kepulauan Riau Health and Stastistic Agency in 2016 the Kepulauan Riau Province had 253,420 km2 consisting of an ocean area of 242,825 km2 (96%) and land area of 10,595.41 km2 (4%). From these data, it is expected that the coastal areas should be able to become a mainstay of the Kepri community's income sources. With the unique potential and economic value, it will be an attraction to disturbances and threats. Therefore, it is necessary to strengthen human resources in coastal areas through education.&#x0D; The life patterns of coastal communities certainly require a different approach to be able to improve education and knowledge of coastal communities, the technology-based solution proposed is The-AT (The Amazing Technology), introduce the maritime region and the potential of the kepulauan riau using visual video and android as a medium to strengthen the knowledge of elementary school children on the coast. The-AT applies games for childrens, which begins with the game showing a map of the area through a gopro camera, then the students are asked to name the island and the location and potential commodities of the area. Android, video and audio will be displayed about the area as information to students who cannot answer correctly. This technology is an effective solution in the formation of thinking patterns for elementary school children regarding the islands in the Kepulauan Riau, so that the formation of a sense of familiarity and a sense to build the potential that exists in the area. At the end it will initiate the technopreneur in the future.

Comparison of still image quality between traditional 35 mm digital and GoPro cameras in a surgical setting

Intraoperative photography is used to obtain images for both education and research purposes, but poses challenges due concerns regarding aseptic technique. Waterproof digital cameras have sterilisable cases that can be used by the surgeon for intraoperative photography. We compared the quality of still intraoperative images obtained by a non-scrubbed observer using a 35 mm single lens reflex (SLR) camera to images obtained by the surgeon using a GoPro camera in a sterilised case. Image quality was scored using a 4 point Likert scale by 3 groups of end users with differing experience: faculty surgeons, surgical residents, and 3rd year veterinary students. Mean ± SEM overall image quality scores were higher for the traditional 35 mm digital SLR camera when compared to the GoPro (3.25 ± 0.08 vs. 2.0 ± 0.08, p < .0001), as were scores for each image characteristic (brightness, colour, sharpness, and contrast). Image quality scores for each camera also differed significantly between user groups, with expert users (faculty and residents) giving lower quality scores when compared to scores from novices (students). Findings suggest that GoPro cameras provide lower intraoperative image quality than digital SLR cameras, although lower quality images may be more accepted by novices than by experienced users.

UAVS ENHANCED NAVIGATION IN OUTDOOR GNSS DENIED ENVIRONMENT USING UWB AND MONOCULAR CAMERA SYSTEMS

Abstract. The demand for small Unmanned Aerial Vehicles (UAVs) is massively increasing these days, due to the wide variety of applications utilizing such vehicles to perform tasks that may be dangerous or just to save time, effort, or cost. Small UAVs navigation system mainly depends on the integration between Global Navigation Satellite Systems (GNSS) and Inertial Measurement Unit (INS) to estimate the Positions, Velocities, and Attitudes (PVT) of the vehicle. Without GNSS such UAVs cannot navigate for long periods of time depending on INS alone, as the low-cost INS typically exhibits massive accumulation of errors during GNSS absence. Given the importance of ensuring full operability of the UAVs even during GNSS signals unavailability, other sensors must be used to bound the INS errors and enhance the navigation system performance. This paper proposes an enhanced UAV navigation system based on integration between monocular camera, Ultra-Wideband (UWB) system, and INS. In addition to using variable EKF weighting scheme. The paper also investigates this integration in the case of low density of UWB anchors, to reduce the cost required for such UWB system infrastructure. A GoPro Camera and UWB rover were attached to the belly of a quadcopter, an on the shelf commercial drone (3DR Solo), during the experimental flight. The velocity of the vehicle is estimated with Optical Flow (OF) from camera successive images, while the range measurements between the UWB rover and the stationary UWB anchors, which were distributed on the field, were used to estimate UAV position.

Preparation and Implementation of a Test Flight of Lightweight, Unmanned Stratospheric Balloon with Gopro Camera Mounted and Analysis of Acquired Material

Abstract Publication contains a description of the preparation and the implementation of a test flight of a stratospheric balloon with a mounted camera GoPro Hero3. Description includes: used equipment, its parameters, role in the success of the mission and the difficulties and limitations that the project team encountered during the preparation and implementation of the flight. The mission was attended by a team of six engineers and scientists from the Remote Sensing Division, who were also involved in the implementation of the HESOFF project. One of the main goals of the HESOFF project was to obtain aerial images on the Krotoszyńska Plate (woj. wielkopolskie) using the Unmanned Aerial Vehicle (UAV) and to carry out remote monitoring of oak stands. The primary goal of an experimental balloon flight was to check the technical operational capability and gain experience in planning and implementing this type of project. During the balloon raising, the video material was acquired in the form of a recording, which later was analyzed. On the basis of the collected information, the conclusions regarding the possibility of implementing a long endurance flight in the stratosphere, illustrating (using a multisensor platform) research surfaces of the HESOFF project were presented. The stages of implementation of the presented mission were divided into following parts: preparation of the flight with the completion of equipment and necessary documents (flight permission), proper flight realization, understood as the release of the balloon and identification of the place where the equipment landed, as well as analysis and presentation of the results.

Markerless 2D kinematic analysis of underwater running: A deep learning approach

Kinematic analysis is often performed with a camera system combined with reflective markers placed over bony landmarks. This method is restrictive (and often expensive), and limits the ability to perform analyses outside of the lab. In the present study, we used a markerless deep learning-based method to perform 2D kinematic analysis of deep water running, a task that poses several challenges to image processing methods. A single GoPro camera recorded sagittal plane lower limb motion. A deep neural network was trained using data from 17 individuals, and then used to predict the locations of markers that approximated joint centres. We found that 300–400 labelled images were sufficient to train the network to be able to position joint markers with an accuracy similar to that of a human labeler (mean difference < 3 pixels, around 1 cm). This level of accuracy is sufficient for many 2D applications, such as sports biomechanics, coaching/training, and rehabilitation. The method was sensitive enough to differentiate between closely-spaced running cadences (45–85 strides per minute in increments of 5). We also found high test–retest reliability of mean stride data, with between-session correlation coefficients of 0.90–0.97. Our approach represents a low-cost, adaptable solution for kinematic analysis, and could easily be modified for use in other movements and settings. Using additional cameras, this approach could also be used to perform 3D analyses. The method presented here may have broad applications in different fields, for example by enabling markerless motion analysis to be performed during rehabilitation, training or even competition environments.

Underwater video surveys provide a more complete picture of littoral fish populations than seine samples in clear Florida springs

Traditional fish-sampling methods may be problematic because of public use or safety concerns. In this study, we compared one common sampling method with video assessment of fish abundance and diversity in three springs that differed in water clarity and structure. At each of four or five sites per spring, we placed one GoPro camera on each bank for 12min and followed the filming with seine sampling. On the video, we counted the maximum number of individuals of each species observed within one frame (MaxN) and summed these counts to produce an estimate of fish abundance (SumMaxN). Then we compared abundance (SumMaxN), species richness and diversity between seine and video samples across all three springs. Video produced higher estimates of abundance (SumMaxN), species richness, and diversity than did seine sampling. However, this effect was largely confined to species richness and diversity differences between sample methods in the structurally complex spring; differences were subtle or non-existent in the low-structure spring and in the turbid spring. In all three springs, video captured relatively more centrarchids; these taxa were captured only rarely in seine samples. Therefore, video sampling performed as well or better than did seine sampling for fish-assemblage assessment in these clear springs.