A Kinect-Based Segmentation of Touching-Pigs for Real-Time Monitoring

Ju, Miso; Choi, Younchang; Seo, Ji Hun; Sa, Jaewon; Lee, Sung-Ju; Chung, Yongwha; Park, Jooyoung

doi:10.3390/s18061746

Cited by 35 publications

(25 citation statements)

References 57 publications

(44 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For measuring the body sizes of pigs, researchers have proposed a portable and automatic measuring system equipped with an ASUS Xtion Pro camera to measure three body dimensions of live pigs (body width, hip width, and body height) with average relative errors within 10.30% [ 42 ]. With the application of a Kinect sensor, aspects of the body measurement of pigs have been investigated, such as automatic recognition of aggressive behavior [ 9 ], real-time monitoring for touching-pigs [ 43 ], live weight determination of pigs from measured dimensions [ 3 , 44 ], normal walking patterns assessment [ 28 ] and so on. For remote measuring with 3D PCD, a series of processing techniques or methods can be employed, such as parameter calibration, Euclidian clustering, RANdom SAmple Consensus (RANSAC) segmentation, viewpoint feature histogram (VFH) extraction, PCD registration, grid reconstruction as well as body measurement.…”

Section: Introductionmentioning

confidence: 99%

Non-Contact Body Measurement for Qinchuan Cattle with LiDAR Sensor

Huang

Zhu

et al. 2018

Sensors

View full text Add to dashboard Cite

The body dimension measurement of large animals plays a significant role in quality improvement and genetic breeding, and the non-contact measurements by computer vision-based remote sensing could represent great progress in the case of dangerous stress responses and time-costing manual measurements. This paper presents a novel approach for three-dimensional digital modeling of live adult Qinchuan cattle for body size measurement. On the basis of capturing the original point data series of live cattle by a Light Detection and Ranging (LiDAR) sensor, the conditional, statistical outliers and voxel grid filtering methods are fused to cancel the background and outliers. After the segmentation of K-means clustering extraction and the RANdom SAmple Consensus (RANSAC) algorithm, the Fast Point Feature Histogram (FPFH) is put forward to get the cattle data automatically. The cattle surface is reconstructed to get the 3D cattle model using fast Iterative Closest Point (ICP) matching with Bi-directional Random K-D Trees and a Greedy Projection Triangulation (GPT) reconstruction method by which the feature points of cattle silhouettes could be clicked and calculated. Finally, the five body parameters (withers height, chest depth, back height, body length, and waist height) are measured in the field and verified within an accuracy of 2 mm and an error close to 2%. The experimental results show that this approach could be considered as a new feasible method towards the non-contact body measurement for large physique livestock.

show abstract

Section: Introductionmentioning

confidence: 99%

Non-Contact Body Measurement for Qinchuan Cattle with LiDAR Sensor

Huang

Zhu

et al. 2018

Sensors

View full text Add to dashboard Cite

show abstract

“…Because the features of an object were much more difficult to identify due to the lower image quality, models trained on this data are at a disadvantage as these are typically what are used in order to create a generalisable model. These factors combined show that the pig dataset was a substantially more difficult dataset than VOC, which is something that previous research has had to overcome by employing post-processing methods [12]. In order to quantify how our implementation performs in these conditions, in addition to testing on the whole test set, we also assessed the detection performance independently for images containing: many pigs, densely packed pigs, overexposed images, and low-light images ( Figure 2).…”

Section: ) Pig Detection Datasetmentioning

confidence: 99%

Automated Individual Pig Localisation, Tracking and Behaviour Metric Extraction Using Deep Learning

2019

View full text Add to dashboard Cite

Individual pig tracking is key to stepping away from group-level treatment and towards individual pig care. By doing so we can monitor individual pig behaviour changes over time and use these as indicators of health and well-being, which, in turn, will assist in the early detection of disease allowing for earlier and more effective intervention. However, it is a much more computationally challenging than performing this task at group level; mistakes in identification and tracking accumulate and, over time, provide noise measures. We combine a deep CNN object localisation method, Faster Region-based convolutional neural network (R-CNN), with two potential real-time multi-object tracking methods in order to create a complete system that can autonomously localise and track individual pigs allowing for the extraction of metrics pertaining to individual pig behaviours from RGB cameras. We evaluate two different transfer learning strategies to adapt Faster R-CNN to our pig detection dataset that is more challenging than conventional tracking benchmark datasets. We are able to localise pigs in individual frames with 0.901 mean average precision (mAP), which then allows us to track individual pigs across video footage with 92% Multi-Object Tracking Accuracy (MOTA) and 73.4% Identity F1-Score (IDF1), and re-identify them after occlusions and dropped frames with 0.862 mAP (0.788 Rank 1 cumulative matching characteristic (CMC)). From these tracks we extract individual behavioural metrics for total distance travelled, time spent idle, and average speed with less than 0.015 mean squared error (MSE) for each. Changes in all these behavioural metrics have value in the detection of pig health and wellbeing.

show abstract

“…First up was the identification of the left-hand gestures. The Kinect sensor was adopted to track the human skeleton [23][24][25]. Second, the left-hand joints were detected.…”

Section: Journal Of Sensorsmentioning

confidence: 99%

The Manipulation of Real-Time Kinect-Based Robotic Arm Using Double-Hand Gestures

et al. 2020

View full text Add to dashboard Cite

This study focused on utilizing the Kinect depth sensor to track double-hand gestures and control a real-time robotic arm. The control system is mainly composed of the microprocessor, a color camera, the depth sensor, and the robotic arm. The Kinect depth sensor was used to take photos of the human body to analyze the skeleton of a human body and obtain the relevant information. Such information was used to identify the gestures of the left hand and the left palm of the user. The gesture of left hand was used as an input command device. The gesture of the right hand was used for imitation movement teaching of robotic arm. From the depth sensor, the real-time images of the human body and the deep information of each joint were collected and converted to the relative positions of the robotic arm. Combining forward kinematics and inverse kinematics and D-H link, the gesture information of the right hand was calculated, which was converted via coordinates into each angle of the motor of the robotic arm. From the color camera, when the left palm was not detected, the user could simply use the right hand to control the action and movement of the real-time robotic arm. When the left palm was detected and 5 fingertips were identified, it meant the start of recording the real-time imitation movement of the robotic arm by the right hand. When 0 fingertip was identified, it meant the stoppage of the above recording. When 2 fingertips were identified, the user could not only control the real-time robotic arm but also repeat the recorded actions.

show abstract

A Kinect-Based Segmentation of Touching-Pigs for Real-Time Monitoring

Cited by 35 publications

References 57 publications

Non-Contact Body Measurement for Qinchuan Cattle with LiDAR Sensor

Non-Contact Body Measurement for Qinchuan Cattle with LiDAR Sensor

Automated Individual Pig Localisation, Tracking and Behaviour Metric Extraction Using Deep Learning

The Manipulation of Real-Time Kinect-Based Robotic Arm Using Double-Hand Gestures

Contact Info

Product

Resources

About