• Title, Summary, Keyword: Mask Video

Search Result 65, Processing Time 0.047 seconds

Video Object Segmentation with Weakly Temporal Information

  • Zhang, Yikun;Yao, Rui;Jiang, Qingnan;Zhang, Changbin;Wang, Shi
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.13 no.3
    • /
    • pp.1434-1449
    • /
    • 2019
  • Video object segmentation is a significant task in computer vision, but its performance is not very satisfactory. A method of video object segmentation using weakly temporal information is presented in this paper. Motivated by the phenomenon in reality that the motion of the object is a continuous and smooth process and the appearance of the object does not change much between adjacent frames in the video sequences, we use a feed-forward architecture with motion estimation to predict the mask of the current frame. We extend an additional mask channel for the previous frame segmentation result. The mask of the previous frame is treated as the input of the expanded channel after processing, and then we extract the temporal feature of the object and fuse it with other feature maps to generate the final mask. In addition, we introduce multi-mask guidance to improve the stability of the model. Moreover, we enhance segmentation performance by further training with the masks already obtained. Experiments show that our method achieves competitive results on DAVIS-2016 on single object segmentation compared to some state-of-the-art algorithms.

An Improved Method for Detection of Moving Objects in Image Sequences Using Statistical Hypothesis Tests

  • Park, Jae-Gark;Kim, Munchurl;Lee, Myoung-Ho;Ahn, Chei-Teuk
    • Proceedings of the Korean Society of Broadcast Engineers Conference
    • /
    • /
    • pp.171-176
    • /
    • 1998
  • This paper resents a spatio-temporal video segmentation method. The algorithm segments each frame of video sequences captured by a static or moving camera into moving objects (foreground) and background using a statistical hypothesis test. In the proposed method, three consecutive image frames are exploited and a hypothesis testing is performed by comparing two means from two consecutive difference images, which results in a T-test. This hypothesis test yields change detection mask that indicates moving areas (foreground) and non-moving areas (background). Moreover, an effective method for extracting object mask form change detection mask is proposed.

  • PDF

Realtime Theft Detection of Registered and Unregistered Objects in Surveillance Video (감시 비디오에서 등록 및 미등록 물체의 실시간 도난 탐지)

  • Park, Hyeseung;Park, Seungchul;Joo, Youngbok
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.24 no.10
    • /
    • pp.1262-1270
    • /
    • 2020
  • Recently, the smart video surveillance research, which has been receiving increasing attention, has mainly focused on the intruder detection and tracking, and abandoned object detection. On the other hand, research on real-time detection of stolen objects is relatively insufficient compared to its importance. Considering various smart surveillance video application environments, this paper presents two different types of stolen object detection algorithms. We first propose an algorithm that detects theft of statically and dynamically registered surveillance objects using a dual background subtraction model. In addition, we propose another algorithm that detects theft of general surveillance objects by applying the dual background subtraction model and Mask R-CNN-based object segmentation technology. The former algorithm can provide economical theft detection service for pre-registered surveillance objects in low computational power environments, and the latter algorithm can be applied to the theft detection of a wider range of general surveillance objects in environments capable of providing sufficient computational power.

A Study on Edge Detection using Directional Mask in Impulse Noise Image (Salt-and-Pepper 잡음 영상에서 방향성 마스크를 이용한 에지 검출에 관한 연구)

  • Lee, Chang-Young;Kim, Nam-Ho
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.18 no.12
    • /
    • pp.2982-2988
    • /
    • 2014
  • The edge detection is a pre-processing of such as image segmentation, image recognition, etc, and many related studies are being conducted both in domestic and abroad. Representative edge detection methods are Sobel, Prewitt, Laplacian, Roberts and Canny edge detectors. Such existing methods are possible for superb detections of edges if edges are detected from videos without noises. However, for video degraded by the salt-and-pepper noise, the edge detection characteristic is shown to be insufficient due to the noise influence. Therefore, in this study, the area is separated as the top, down, left and right from the mask's center pixel first to acquire a superb edge detection characteristic from the video damaged by the salt-and-pepper noise. And the algorithm that detects the final edge by applying the directional mask on the assumed factor of mask that is obtained according to the result of determination for the noise status of representative pixel value of each area.

Awake intubation in a patient with huge orocutaneous fistula: a case report

  • Kim, Hye-Jin;Kim, So-Hyun;Kim, Tae-Heung;Yoon, Ji-Young;Kim, Cheul-Hong;Kim, Eun-Jung
    • Journal of Dental Anesthesia and Pain Medicine
    • /
    • v.17 no.4
    • /
    • pp.313-316
    • /
    • 2017
  • Mask ventilation, the first step in airway management, is a rescue technique when endotracheal intubation fails. Therefore, ordinary airway management for the induction of general anesthesia cannot be conducted in the situation of difficult mask ventilation (DMV). Here, we report a case of awake intubation in a patient with a huge orocutaneous fistula. A 58-year-old woman was scheduled to undergo a wide excision, reconstruction with a reconstruction plate, and supraomohyoid neck dissection on the left side and an anterolateral thigh flap due to a huge orocutaneous fistula that occurred after a previous mandibulectomy and flap surgery. During induction, DMV was predicted, and we planned an awake intubation. The patient was sedated with dexmedetomidine and remifentanil. She was intubated with a nasotracheal tube using a video laryngoscope, and spontaneous ventilation was maintained. This case demonstrates that awake intubation using a video laryngoscope can be as good as a fiberoptic scope.

Mask Wearing Detection System using Deep Learning (딥러닝을 이용한 마스크 착용 여부 검사 시스템)

  • Nam, Chung-hyeon;Nam, Eun-jeong;Jang, Kyung-Sik
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.25 no.1
    • /
    • pp.44-49
    • /
    • 2021
  • Recently, due to COVID-19, studies have been popularly worked to apply neural network to mask wearing automatic detection system. For applying neural networks, the 1-stage detection or 2-stage detection methods are used, and if data are not sufficiently collected, the pretrained neural network models are studied by applying fine-tuning techniques. In this paper, the system is consisted of 2-stage detection method that contain MTCNN model for face recognition and ResNet model for mask detection. The mask detector was experimented by applying five ResNet models to improve accuracy and fps in various environments. Training data used 17,217 images that collected using web crawler, and for inference, we used 1,913 images and two one-minute videos respectively. The experiment showed a high accuracy of 96.39% for images and 92.98% for video, and the speed of inference for video was 10.78fps.

Use of laryngeal mask after repeated endotracheal intubation failure in a patient with tracheobronchopathia osteochondroplastica: case report

  • Kim, Sang Gyun;Kim, Hyun;Son, Jong Chul;Lee, Ji-Hyang;An, Jihyun;Kim, Eunju
    • Kosin Medical Journal
    • /
    • v.33 no.2
    • /
    • pp.252-256
    • /
    • 2018
  • We report a case of difficult endotracheal intubation in a patient with tracheobronchopathia osteochondroplastica. A 65-year-old man was scheduled to undergo ulnar nerve decompression and ganglion excisional biopsy under general anesthesia. During induction of general anesthesia, an endotracheal tube could not be advanced through the vocal cords due to resistance. A large number of nodules were identified below the vocal cords using a $Glidescope^{(R)}$ video-laryngoscopy, and fiberoptic bronchoscopy revealed irregular nodules on the surface of the entire trachea and the main bronchus below the vocal cords. Use of a small endotracheal tube was attempted and failed. a laryngeal mask airway (LMA $Supreme^{TM}$) rather than further intubation was successfully used to maintain the airway.

Shot Change Detection Using Fuzzy Clustering Method on MPEG Video Frames (퍼지 클러스터링 기법을 이용한 MPEG 비디오의 장면 전환 검출)

  • Lim, Seong-Jae;Kim, Woon;Lee, Bae-Ho
    • Proceedings of the IEEK Conference
    • /
    • /
    • pp.159-162
    • /
    • 2000
  • In this paper, we propose an efficient method to detect shot changes in compressed MPEG video data by using reference features among video frames. The reference features among video frames imply the similarities among adjacent frames by prediction coded type of each frame. A shot change is detected if the similarity degrees of a frame and its adjacent frames are low. And the shot change detection algorithm is improved by using Fuzzy c-means (FCM) clustering algorithm. The FCM clustering algorithm uses the shot change probabilities evaluated in the mask matching of reference ratios and difference measure values based on frame reference ratios.

  • PDF

Efficient Shot Change Detection Using Clustering Method on MPEG Video Frames (MPEG 비디오 프레임에서 FCM 클러스터링 기법을 이용한 효과적인 장면 전환 검출)

  • Lim, Seong-Jae;Lee, Bae-Ho
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • /
    • pp.751-754
    • /
    • 2000
  • In this paper, we propose an efficient method to detect abrupt shot changes in compressed MPEG video data by using reference ratios among video frames. The reference ratios among video frames imply the degree of similarities among adjacent frames by prediction coded type of each frames. A shot change is detected if the similarity degrees of a frame and its adjacent frames are low. This paper proposes an efficient shot change detection algorithm by using Fuzzy c-means(FCM) clustering algorithm. The FCM clustering uses the shot change probabilities evaluated in the mask matching of reference ratios and difference measure values based on frame reference ratios.

  • PDF

A Study on Effective Stitching Technique of 360° Camera Image (360° 카메라 영상의 효율적인 스티칭 기법에 관한 연구)

  • Lee, Lang-Goo;Chung, Jean-Hun
    • Journal of Digital Convergence
    • /
    • v.16 no.2
    • /
    • pp.335-341
    • /
    • 2018
  • This study is a study on effective stitching technique for video recorded by using a dual-lens $360^{\circ}$ camera composed of two fisheye lenses. First of all, this study located a problem in the result of stitching by using a bundled program. And the study was carried out, focusing on looking for a stitching technique more efficient and closer to perfect by comparatively analyzing the results of stitching by using Autopano Video Pro and Autopano Giga, professional stitching program. As a result, it was shown that the problems of bundled program were horizontal and vertical distortion, exposure and color mismatch and unsmooth stitching line. And it was possible to solve the problem of the horizontal and vertical by using Automatic Horizon and Verticals Tool of Autopano Video Pro and Autopano Giga, problem of exposure and color by using Levels, Color and Edit Color Anchors and problem of stitching line by using Mask function. Based on this study, it is to be hoped that $360^{\circ}$ VR video content closer to perfect can be produced by efficient stitching technique for video recorded by using dual-lens $360^{\circ}$ camera in the future.