Recognition Of Body Parts Research Articles

Selection of hand features based on Random Forest algorithm and hand shape recognition

To obtain effective features applicable to hand morphology recognition, the method of obtaining effective feature indicators based on the Random Forest (RF) algorithm to downscale hand morphology parameters is proposed. Firstly, 232 female university students collected three-dimensional hand information, constructed auxiliary point, line, and surface standardised measurement methods, obtained 33 characteristic parts of human dimensions, and used k-means clustering for hand morphology subdivision. Hand morphology can be divided into three categories: short, slender and broad. The RF algorithm is used for feature index importance assessment and hand shape recognition model. The accuracy of the feature metrics determined by the RF algorithm, PCA, and VC method applied to hand shape recognition is compared and analysed to verify the effectiveness of the dimensionality reduction of the RF algorithm. The results showed that the feature indexes used for hand shape recognition were five items: hand length, thickness at the metacarpal, thenar width, the distance between the thumb and index finger, and distance from the root of the little finger to the centre of the wrist. Using the RF algorithm to reduce the dimensionality is more effective; the average recognition accuracy of the four hand shape recognition models is 93.78% on average, compared with PCA and VC reduction methods, the average accuracy of hand shape recognition models is increased by 19.17%, and 14.86% respectively. The study's results can provide methodological references for the objective selection of characteristic indicators and morphological recognition of human body parts.

Pengaruh Video Animasi terhadap Kemampuan Mengenal Anggota Tubuh pada Anak

Learning about body recognition is still minimal, children still do not understand which body parts can be touched and which others cannot touch. The aim of the study was to find out how animated videos affect the ability to recognize body parts in children aged 5–6 years at Beringin Kindergarten. The research method is quantitative, using video animation experiments and pre- and post-test research designs, the research subjects are 20 children aged 5–6 years, data collection techniques are observation, documentation. The results showed that there was an effect on the recognition of limbs before and after using video animation media, as evidenced by the total value obtained by the pretest, which was 11.7, after 8 treatments, there was an increase and change of 27.55; and the posttest obtained a value of 30.6. This proves a significant difference before and after giving treatment through animated video media for the introduction of body parts to children. The conclusion is that the recognition of body parts through the use of animated video media has an effect on the ability to recognize body parts

Recognition and 3D Visualization of Human Body Parts and Bone Areas Using CT Images

Abstract The advent of medical imaging significantly assisted in disease diagnosis and treatment. This study introduces to a framework for detecting several human body parts in Computerised Tomography (CT) images formatted in DICOM files. In addition, the method can highlight the bone areas inside CT images and transform 2D slices into a visual 3D model to illustrate the structure of human body parts. Firstly, we leveraged shallow convolutional Neural Networks to classify body parts and detect bone areas in each part. Then, Grad-CAM was applied to highlight the bone areas. Finally, Insight and Visualization libraries were utilized to visualize slides in 3D of a body part. As a result, the classifiers achieved 98 % in F1-score in the classification of human body parts on a CT image dataset, including 1234 slides capturing body parts from a woman for the training phase and 1245 images from a male for testing. In addition, distinguishing between bone and non-bone images can reach 97 % in F1-score on the dataset generated by setting a threshold value to reveal bone areas in CT images. Moreover, the Grad-CAM-based approach can provide clear, accurate visualizations with segmented bones in the image. Also, we successfully converted 2D slice images of a body part into a lively 3D model that provided a more intuitive view from any angle. The proposed approach is expected to provide an interesting visual tool for supporting doctors in medical image-based disease diagnosis.

Modelling Proper and Improper Sitting Posture of Computer Users Using Machine Vision for a Human–Computer Intelligent Interactive System during COVID-19

Human posture recognition is one of the most challenging tasks due to the variation in human appearance, changes in the background and illumination, additional noise in the frame, and diverse characteristics and amount of data generated. Aside from these, generating a high configuration for recognition of human body parts, occlusion, nearly identical parts of the body, variations of colors due to clothing, and other various factors make this task one of the hardest in computer vision. Therefore, these studies require high-computing devices and machines that could handle the computational load of this task. This study used a small-scale convolutional neural network and a smartphone built-in camera to recognize proper and improper sitting posture in a work-from-home setup. Aside from the recognition of body points, this study also utilized points’ distances and angles to help in recognition. Overall, the study was able to develop two objective datasets capturing the left and right side of the participants with the supervision and guidance of licensed physical therapists. The study shows accuracies of 85.18% and 92.07%, and kappas of 0.691 and 0.838, respectively. The system was developed, implemented, and tested in a work-from-home environment.

Recognition of animal body parts via supervised learning

The application of machine learning techniques has facilitated the automatic annotation of behavior in video sequences, offering a promising approach for ethological studies by reducing the manual effort required for annotating each video frame. Nevertheless, before solely relying on machine-generated annotations, it is essential to evaluate the accuracy of these annotations to ensure their reliability and applicability. While it is conventionally accepted that there cannot be a perfect annotation, the degree of error associated with machine-generated annotations should be commensurate with the error between different human annotators. The authors hypothesized that machine learning supervised with adequate human annotations would be able to accurately predict body parts from video sequences. The authors conducted a comparative analysis of the quality of annotations generated by humans and machines for the body parts of sheep during treadmill walking. For human annotation, two annotators manually labeled six body parts of sheep in 300 frames. To generate machine annotations, the authors employed the state-of-the-art pose-estimating library, DeepLabCut, which was trained using the frames annotated by human annotators. As expected, the human annotations demonstrated high consistency between annotators. Notably, the machine learning algorithm also generated accurate predictions, with errors comparable to those between humans. They also observed that abnormal annotations with a high error could be revised by introducing Kalman Filtering, which interpolates the trajectory of body parts over the time series, enhancing robustness. Their results suggest that conventional transfer learning methods can generate behavior annotations as accurate as those made by humans, presenting great potential for further research.

Visual recognition of single body parts in natural images

Visual recognition of single body parts in natural images

Objects with motor valence affect the visual processing of human body parts: Evidence from behavioural and ERP studies

Recent findings indicate that the mental representation of an object contains crucial information about the motor interactions relevant for its intended functional use, suggesting a possible action-specific link with body effectors. For example, in the visual system, the extrastriate body area (EBA) responds to full body and body part images according to a functional/semantic organizational principle. However, the pliancy of the relationship between objects and body parts remains under-investigated. The present study aims to i) investigate this relationship more directly by assessing whether recognition of specific body parts can be facilitated by a brief exposure to functionally-related objects (Experiment 1) and ii) whether the functional relationship between objects and body parts modulates a posterior body-sensitive ERP waveform, peaking around 200 msec, and the more centro-parietal P300, linked to item categorization processes and visual awareness (Experiment 2). Participants were asked to quickly recognize targets (pictures of hands or feet) preceded by a functionally related (e.g., drum for hand target), unrelated (e.g., drum for foot target), or neutral (e.g., unknown object for both targets) prime. Findings showed that participants’ performance was significantly more accurate with related than unrelated primes and that ERP amplitudes were modulated by the relationship between the prime and the target. These findings confirm the existence of action-specific links between objects and body parts and expand on recent findings on categorical organization of neural responses to human effectors in the visual system.

Deep learning-based body part recognition algorithm for three-dimensional medical images.

The automatic recognition of human body parts in three-dimensional medical images is important in many clinical applications. However, methods presented in prior studies have mainly classified each two-dimensional (2D) slice independently rather than recognizing a batch of consecutive slices as a specific body part. In this study, we aim to develop a deep learning-based method designed to automatically divide computed tomography (CT) and magnetic resonance imaging (MRI) scans into five consecutive body parts: head, neck, chest, abdomen, and pelvis. A deep learning framework was developed to recognize body parts in two stages. In the first preclassification stage, a convolutional neural network (CNN) using the GoogLeNet Inception v3 architecture and a long short-term memory (LSTM) network were combined to classify each 2D slice; the CNN extracted information from a single slice, whereas the LSTM employed rich contextual information among consecutive slices. In the second postprocessing stage, the input scan was further partitioned into consecutive body parts by identifying the optimal boundaries between them based on the slice classification results of the first stage. To evaluate the performance of the proposed method, 662 CT and 1434 MRI scans were used. Our method achieved a very good performance in 2D slice classification compared with state-of-the-art methods, with overall classification accuracies of 97.3% and 98.2% for CT and MRI scans, respectively. Moreover, our method further divided whole scans into consecutive body parts with mean boundary errors of 8.9 and 3.5mm for CT and MRI data, respectively. The proposed method significantly improved the slice classification accuracy compared with state-of-the-art methods, and further accurately divided CT and MRI scans into consecutive body parts based on the results of slice classification. The developed method can be employed as an important step in various computer-aided diagnosis and medical image analysis schemes.

Use of recognition of laterality through implicit motor imagery for the improvement of postural control and balance in subacute stroke patients: a randomized controlled study.

Motor Imagery techniques may be used as a complement to the recovery of motor sequelae after a stroke, as during the evocation of a movement the activation of neuronal circuits involved in the actual execution of the movement occurs. A simple-blind randomized controlled trial was conducted. A total of 38 patients were randomly assigned to a study group. Both groups performed, for four weeks, five weekly sessions of neurorehabilitation and three weekly sessions of experimental or control intervention, respectively.The experimental group training the recognition of laterality, while the control group the recognition of body parts. Participants were evaluated pre and post intervention with posturography parameters -Sway area (AREA), Sway path length (LONG), difference in weigthload between lower limbs (DIFLOAD)-, the Berg Balance scale (BBS), the Barthel Index (BI), the Time Up and Go Test (TUG), the Functional Ambulation Categories (FAC), and the quality-of-life scale for stroke (ECVI-38). After performing the intragroup analysis, statistical significance was obtained for AREA (p < 0.001), LONG (p = 0.04), DIFLOAD (p = 0.02), BBS (p < 0.001), BI (p < 0.001), FAC (p < 0.001), and ECVI-38 (p < 0.001) in the experimental group; and for DIFLOAD (p = 0.01), BBS (p = 0.001), BI (p = 0.001), TUG (p = 0.04), FAC (p = 0.03), and ECVI-38 (p = 0.003) in the control group. In the intergroup analysis, statistical significance was obtained for AREA (p = 0.03), BBS (p = 0.03), FAC (p = 0.02) and ECVI-38 (p = 0.002) at postintervention time. Combined use of physical rehabilitation and recognition of laterality through implicit motor imagery tasks, improves balance and functions related to postural control in subacute stroke patients.

Determination of Body Parts in Holstein Friesian Cows Comparing Neural Networks and k Nearest Neighbour Classification.

Simple SummaryThe recognition of objects in three dimensional (3D) data is challenging, especially when it comes to the partition of objects into predefined segments. In this study, two machine learning approaches were applied to recognize the body parts head, rump, back, legs and udder of dairy cows in 3D data recorded with Microsoft Kinect V1, for usage in automated exterior evaluation. Five properties of data points were used as features in order to train a k nearest neighbour classifier and a neural network to determine the body part on which the point was located. Both algorithms used for this feature-based approach are computationally faster and more flexible compared to a model-based object detection priorly used for the same purpose. Both methods could be considered successful for the determination of body parts via pixel properties. Reaching high to very high overall accuracies and very small Hamming losses, the k nearest neighbour classification is superior to the neural network which reached medium to high evaluation metrics. However, k nearest neighbour classification is at runtime prone to higher costs regarding computational time and memory, while once trained, the neural network delivers the classification results very quickly. This needs also to be taken into account in the decision of which method should be implemented.Machine learning methods have become increasingly important in animal science, and the success of an automated application using machine learning often depends on the right choice of method for the respective problem and data set. The recognition of objects in 3D data is still a widely studied topic and especially challenging when it comes to the partition of objects into predefined segments. In this study, two machine learning approaches were utilized for the recognition of body parts of dairy cows from 3D point clouds, i.e., sets of data points in space. The low cost off-the-shelf depth sensor Microsoft Kinect V1 has been used in various studies related to dairy cows. The 3D data were gathered from a multi-Kinect recording unit which was designed to record Holstein Friesian cows from both sides in free walking from three different camera positions. For the determination of the body parts head, rump, back, legs and udder, five properties of the pixels in the depth maps (row index, column index, depth value, variance, mean curvature) were used as features in the training data set. For each camera positions, a k nearest neighbour classifier and a neural network were trained and compared afterwards. Both methods showed small Hamming losses (between 0.007 and 0.027 for k nearest neighbour (kNN) classification and between 0.045 and 0.079 for neural networks) and could be considered successful regarding the classification of pixel to body parts. However, the kNN classifier was superior, reaching overall accuracies 0.888 to 0.976 varying with the camera position. Precision and recall values associated with individual body parts ranged from 0.84 to 1 and from 0.83 to 1, respectively. Once trained, kNN classification is at runtime prone to higher costs in terms of computational time and memory compared to the neural networks. The cost vs. accuracy ratio for each methodology needs to be taken into account in the decision of which method should be implemented in the application.

Visual Hand Recognition in Hand Laterality and Self-Other Discrimination Tasks: Relationships to Autistic Traits and Positive Body Image.

In a study concerning visual body part recognition, a “self-advantage” effect, whereby self-related body stimuli are processed faster and more accurately than other-related body stimuli, was revealed, and the emergence of this effect is assumed to be tightly linked to implicit motor simulation, which is activated when performing a hand laterality judgment task in which hand ownership is not explicitly required. Here, we ran two visual hand recognition tasks, namely, a hand laterality judgment task and a self-other discrimination task, to investigate (i) whether the self-advantage emerged even if implicit motor imagery was assumed to be working less efficiently and (ii) how individual traits [such as autistic traits and the extent of positive self-body image, as assessed via the Autism Spectrum Quotient (AQ) and the Body Appreciation Scale-2 (BAS-2), respectively] modulate performance in these hand recognition tasks. Participants were presented with hand images in two orientations [i.e., upright (egocentric) and upside-down (allocentric)] and asked to judge whether it was a left or right hand (an implicit hand laterality judgment task). They were also asked to determine whether it was their own, or another person’s hand (an explicit self-other discrimination task). Data collected from men and women were analyzed separately. The self-advantage effect in the hand laterality judgment task was not revealed, suggesting that only two orientation conditions are not enough to trigger this motor simulation. Furthermore, the men’s group showed a significant positive correlation between AQ scores and reaction times (RTs) in the laterality judgment task, while the women’s group showed a significant negative correlation between AQ scores and differences in RTs and a significant positive correlation between BAS-2 scores and dprime in the self-other discrimination task. These results suggest that men and women differentially adopt specific strategies and/or execution processes for implicit and explicit hand recognition tasks.

Determination of Lying Posture through Recognition of Multitier Body Parts

In this paper, we propose a lying posture discrimination algorithm to monitor the behavior pattern of a person lying in bed. Using FSR (force sensing resistor) sensors arranged in a grid structure, three major body parts such as head, shoulder, and hips are identified, and six lying positions are determined based on these body parts. Head, shoulder, and hips are relatively high in pressure and are easy to distinguish due to low movement. In addition, we have effectively limited the search space to identify the above body parts based on standard body dimensions. Experimental results show that there is a correlation between pressure distribution and lying posture. The results of this study can be applied to the monitoring of sleep posture and can also be used as an aid to prevent fall accidents and the occurrence of pressure ulcers of elderly patients who are unable to move.

The Impact of an Educational Media Intervention to Support Children’s Early Learning in Rwanda

Children in developing countries often lack sufficient support for early learning skills prior to beginning school. This research evaluates an educational media intervention using an animated cartoon program, Akili and Me. The program was originally created in Tanzania to teach early learning skills. This program was adapted in content and language use in this study in Rwanda. The two-week intervention involved primary school students (mean age = 7.1 years) who were randomized into two groups (intervention and comparison group). The intervention group viewed one Akili and Me episode a day for five days. This viewing was repeated the following week. Similarly, the comparison watched the same amount of television but the content consisted of local popular programs. Baseline and follow-up assessments evaluated 10 areas of early learning, using an adaptation of the International Development and Early Learning Assessment (IDELA), and also children’s media receptivity. At follow-up, children in the intervention program, Akili and Me, had significantly higher scores for counting, number recognition, shape knowledge, letter identification, color identification, body part recognition, health knowledge, and vocabulary. The analyses provide promising evidence that locally produced educational media interventions can impact early learning skills, even among children living in resource-poor communities.

Human Pose Estimation in Images and Videos

Estimation of human poses is an interesting and challenging topic in the field of Computer vision. It includes some un-noticed challenges like background effect, the color of the dress, skin tones and many other unpredictable challenges. This is a workable concept because it can be used in sign language recognition, correlating various pose styles from different parts of the world and in medical applications. A deep structure which can represent a man’s body in different models will help in improved recognition of body parts and the spatial correlation between them. For hand detection, features based on hand shape and representation of geometrical details are derived with the help of hand contour. An adaptive and unsupervised approach based on Voronoi region is primarily used for the color image segmentation problem. This process includes identification of key points of the body, which may include body joints and parts. The identification parts will be tough due to small joints and occlusions. Identification of Image features is described in this paper with the help of Box Model Based Estimation, Speed up robust features and finally with Optical flow tracking algorithm. In Optical flow tracking algorithm, we have used Horn-Schunk algorithm to determine featural changes in the images.

The experiences and beliefs of patients with complex regional pain syndrome: An exploratory survey study

Objectives To determine the beliefs and describe the health care experiences of patients with complex regional pain syndrome. Methods A survey tool for patients with complex regional pain syndrome was designed for this study. The survey tool collected self-reported measures associated with pain, disability, health care experiences, education, beliefs, and treatments. Results Thirty-one patients attending physical therapy for complex regional pain syndrome (mean age 40.48; female n = 20) completed the survey. Patients with presented with high levels of pain and disability and reported various changes associated with altered neuroplasticity such as confused body part recognition, left/right discrimination, neglect, and spreading pain. The patients' experiences with diagnostic testing and interventions are not in line with the current pain science research and/or evidence-based practice. Overall, patients are ill-informed, confused, and receive conflicting information. Discussion The suffering associated with complex regional pain syndrome is real, as told by patients. This suffering coincides with a lack of consensus by health care providers and conflicting information on complex regional pain syndrome. Overall, patient experiences show health care providers are not up to date with the current best-evidence regarding complex regional pain syndrome.

Познавање на делови на тело кај лица со интелектуална попреченост

Познавање на делови на тело кај лица со интелектуална попреченост

Supervised framework for automatic recognition and retrieval of interaction: a framework for classification and retrieving videos with similar human interactions

This study presents supervised framework for automatic recognition and retrieval of interactions (SAFARRIs), a supervised learning framework to recognise interactions such as pushing, punching, and hugging, between a pair of human performers in a video shot. The primary contribution of the study is to extend the vectors of locally aggregated descriptors (VLADs) as a compact and discriminative video encoding representation, to solve the complex class partitioning problem of recognising human interaction. An initial codebook is generated from the training set of video shots, by extracting feature descriptors around the spatiotemporal interest points computed across frames. A bag of action words is generated by encoding the first‐order statistics of the visual words using VLAD. Support vector machine classifiers (1 against all) are trained using these codebooks. The authors have verified SAFARRI's accuracy for classification and retrieval (query by example). SAFARRI is free from tracking or recognition of body parts and capable of identifying the region of interaction in video shots. It gives superior retrieval and classification performances over recently proposed methods, on two publicly available human interaction datasets.

Holistic processing for bodies and body parts: New evidence from stereoscopic depth manipulations.

Although holistic processing has been documented extensively for upright faces, it is unclear whether it occurs for other visual categories with more extensive substructure, such as body postures. Like faces, body postures have high social relevance, but they differ in having fine-grain organization not only of basic parts (e.g., arm) but also subparts (e.g., elbow, wrist, hand). To compare holistic processing for whole bodies and body parts, we employed a novel stereoscopic depth manipulation that creates either the percept of a whole body occluded by a set of bars, or of segments of a body floating in front of a background. Despite sharing low-level visual properties, only the stimulus perceived as being behind bars should be holistically "filled in" via amodal completion. In two experiments, we tested for better identification of individual body parts within the context of a body versus in isolation. Consistent with previous findings, recognition of body parts was better in the context of a whole body when the body was amodally completed behind occluders. However, when the same bodies were perceived as floating in strips, performance was significantly worse, and not significantly different, from that for amodally completed parts, supporting holistic processing of body postures. Intriguingly, performance was worst for parts in the frontal depth condition, suggesting that these effects may extend from gross body organization to a more local level. These results provide suggestive evidence that holistic representations may not be "all-or-none," but rather also operate on body regions of more limited spatial extent.

Multi-Instance Deep Learning: Discover Discriminative Local Anatomies for Bodypart Recognition.

In general image recognition problems, discriminative information often lies in local image patches. For example, most human identity information exists in the image patches containing human faces. The same situation stays in medical images as well. "Bodypart identity" of a transversal slice-which bodypart the slice comes from-is often indicated by local image information, e.g., a cardiac slice and an aorta arch slice are only differentiated by the mediastinum region. In this work, we design a multi-stage deep learning framework for image classification and apply it on bodypart recognition. Specifically, the proposed framework aims at: 1) discover the local regions that are discriminative and non-informative to the image classification problem, and 2) learn a image-level classifier based on these local regions. We achieve these two tasks by the two stages of learning scheme, respectively. In the pre-train stage, a convolutional neural network (CNN) is learned in a multi-instance learning fashion to extract the most discriminative and and non-informative local patches from the training slices. In the boosting stage, the pre-learned CNN is further boosted by these local patches for image classification. The CNN learned by exploiting the discriminative local appearances becomes more accurate than those learned from global image context. The key hallmark of our method is that it automatically discovers the discriminative and non-informative local patches through multi-instance deep learning. Thus, no manual annotation is required. Our method is validated on a synthetic dataset and a large scale CT dataset. It achieves better performances than state-of-the-art approaches, including the standard deep CNN.

Application of AdaBoost Algorithm in Basketball Player Detection

Video materials contain huge amount of information. Their storage in databases and analysis by various algorithms is a constantly developing area. This paper presents the process of basketball game analysis by AdaBoost algorithm. This algorithm is mainly used for face and body parts recognition, and was not tested on player detection in basketball. It consists of a linear combination of weak classifiers. In this paper, we used stumps, i.e. decision trees with only one level as such classifiers. The aim of this research is to assess the accuracy of this algorithm when applied in player detection during basketball games. We examined the capabilities of AdaBoost algorithm on a video footage obtained from the single moving camera, without any previous processing. First training was performed using images of a basketball player's entire body (head, legs, arms and torso), while the second training was performed using images of a head and torso. By applying the algorithm to the given set of images that include head and torso, the algorithm obtained an accuracy of 70.5%. Training on the set of entire body images was not successful due to the large amount of background that goes into the training, and which represents noise in training process. This research concluded that AdaBoost could not be applied to object detection in sports events. We also concluded that this algorithm gives much better results when applied on simpler objects (like face recognition) and that its application could be in detection of players' body parts or as a first step in object detection in order to eliminate as much area as possible. Its application in detecting players' upper body or entire players gives large number of false positive, which makes algorithm inapplicable in real situations.