• Journal of Animal Science and Technology
• /
• v.66 no.1
• /
• pp.167-177
• /
• 2024

Iris segmentation is an initial step for identifying the biometrics of animals when establishing a traceability system for livestock. In this study, we propose a deep learning framework for pixel-wise segmentation of bovine iris with a minimized use of annotation labels utilizing the BovineAAEyes80 public dataset. The proposed image segmentation framework encompasses data collection, data preparation, data augmentation selection, training of 15 deep neural network (DNN) models with varying encoder backbones and segmentation decoder DNNs, and evaluation of the models using multiple metrics and graphical segmentation results. This framework aims to provide comprehensive and in-depth information on each model's training and testing outcomes to optimize bovine iris segmentation performance. In the experiment, U-Net with a VGG16 backbone was identified as the optimal combination of encoder and decoder models for the dataset, achieving an accuracy and dice coefficient score of 99.50% and 98.35%, respectively. Notably, the selected model accurately segmented even corrupted images without proper annotation data. This study contributes to the advancement of iris segmentation and the establishment of a reliable DNN training framework.

• Journal of Broadcast Engineering
• /
• v.16 no.1
• /
• pp.44-53
• /
• 2011

In this paper, we propose a 3D conversion algorithm of 2D video using depth layer partition method. In the proposed algorithm, we first set frame groups using cut detection algorithm. Each divided frame groups will reduce the possibility of error propagation in the process of motion estimation. Depth image generation is the core technique in 2D/3D conversion algorithm. Therefore, we use two depth map generation algorithms. In the first, segmentation and motion information are used, and in the other, edge directional histogram is used. After applying depth layer partition algorithm which separates objects(foreground) and the background from the original image, the extracted two depth maps are properly merged. Through experiments, we verify that the proposed algorithm generates reliable depth map and good conversion results.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.12
• /
• pp.1919-1925
• /
• 2022

In this paper, we propose an image-to-image translation approach based on the conditional generative adversarial network for semantic segmentation. Semantic segmentation is the task of clustering parts of an image together which belong to the same object class. Unlike the traditional pixel-wise classification approach, the proposed method parses an input RGB image to its corresponding semantic segmentation mask using a pixel regression approach. The proposed method is based on the Pix2Pix image synthesis method. We employ residual connections-based convolutional neural network architectures for both the generator and discriminator architectures, as the residual connections speed up the training process and generate more accurate results. The proposed method has been trained and tested on the NYU-depthV2 dataset and could achieve a good mIOU value (49.5%). We also compare the proposed approach to the current methods in semantic segmentation showing that the proposed method outperforms most of those methods.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.1
• /
• pp.234-240
• /
• 2016

We propose a novel framework for markerless camera-based games. By using a visual camera, our method may yield robust human body segmentation with high performance comparable to the segmentation using depth cameras. The edges of human bodies are detected by subtracting gradient backgrounds, and human body regions are segmented by the operations based on mathematical morphology. Collisions between detected regions and virtual objects are determined by finding the colliding time using intra-frame positions of virtual objects. Experimental results show that the proposed method may produce robust segmentation of human bodies, thereby and the collision responses are more accurate than previous methods. Therefore, the proposed framework can be widely used in camera-based games requiring high performance.

• Journal of Broadcast Engineering
• /
• v.27 no.7
• /
• pp.1011-1020
• /
• 2022

Recent high-performance panoptic segmentation models are based on transformer architectures. However, transformer-based panoptic segmentation methods are basically slower than convolution-based methods, since the attention mechanism in the transformer requires quadratic complexity w.r.t. image resolution. Also, sine and cosine computation for positional embedding in the transformer also yields a bottleneck for computation time. To address these problems, we adopt three modules to speed up the inference runtime of the transformer-based panoptic segmentation. First, we perform channel-level reduction using depth-wise separable convolution for inputs of the transformer decoder. Second, we replace sine and cosine-based positional encoding with convolution operations, called conv-embedding. We also apply a separable self-attention to the transformer encoder to lower quadratic complexity to linear one for numbers of image pixels. As result, the proposed model achieves 44% faster frame per second than baseline on ADE20K panoptic validation dataset, when we use all three modules.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.12C
• /
• pp.1147-1153
• /
• 2009

We propose a novel image segmentation and segment-based stereo matching technique using color, depth, and motion information. Proposed technique firstly splits reference images into foreground region or background region using depth information from depth camera. Then each region is segmented into small segments with color information. Moreover, extracted segments in current frame are tracked in the next frame in order to maintain depth consistency between frames. The initial depth from the depth camera is also used to set the depth search range for stereo matching. Proposed segment-based stereo matching technique was compared with conventional one without foreground and background separation and other conventional one without motion tracking of segments. Simulation results showed that the improvement of segment extraction and depth estimation consistencies by proposed technique compared to conventional ones especially at the static background region.

• Journal of Internet Computing and Services
• /
• v.18 no.4
• /
• pp.61-67
• /
• 2017

Recently, CCTV has been combined with areas such as big data, artificial intelligence, and image analysis to detect various abnormal behaviors and to detect and analyze the overall situation of objects such as people. Image analysis research for this intelligent video surveillance function is progressing actively. However, CCTV images using 2D information generally have limitations such as object misrecognition due to lack of topological information. This problem can be solved by adding the depth information of the object created by using two cameras to the image. In this paper, we perform background modeling using Mixture of Gaussian technique and detect whether there are moving objects by segmenting the foreground from the modeled background. In order to perform the depth information-based segmentation using the RGB information-based segmentation results, stereo-based depth maps are generated using two cameras. Next, the RGB-based segmented region is set as a domain for extracting depth information, and depth-based segmentation is performed within the domain. In order to detect the center point of a robustly segmented object and to track the direction, the movement of the object is tracked by applying the CAMShift technique, which is the most basic object tracking method. From the experiments, we prove the efficiency of the proposed object detection and tracking method using the RGB-D model.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.639-640
• /
• 2013

In this study, a new method that finds an area where interesting objects are placed to generate DIBR-based intermediate images with higher quality. This method complements the existing object segmentation algorithm called Grabcut by finding the bounding box automatically, whereas the existing algorithm requires a user to select the region specifically. Then, the histogram of the depth map information is then used to separate the background and the frontal objects after applying the GrabCut algorithm. By using the new method, it is found that it produces better result than the existing algorithm. This paper describes the new method and future research.

• Korean Journal of Remote Sensing
• /
• v.13 no.2
• /
• pp.133-150
• /
• 1997

The purpose of this paper is to develop an algorithm of classification and interpretation of seafloor based on side scan sonar data. The algorithm consists of mosaicking of sonar data using navigation data, correction and compensation of the acouctic amplitude data considering the charateristics of the side scan sonar system, and segmentation of the seafloor using digital image processing techniques. The correction and compensation process is essential because there is usually difference in acoustic amplitudes from the same distance of the port-side and the starboard-side and the amplitudes become attenuated as the distance is increasing. In this paper, proposed is an algorithm of compensating the side scan sonar data, and its result is compared with the mosaicking result without any compensation. The algorithm considers the amplitude characteristics according to the tow-fish's depth as well as the attenuation trend of the side scan sonar along the beam positions. This paper also proposes an image segmentation algorithm based on the texture, where the criterion is the maximum occurence related with gray level. The preliminary experiment has been carried out with the side scan sonar data and its result is demonstrated.

• ETRI Journal
• /
• v.35 no.3
• /
• pp.491-500
• /
• 2013

A connected component analysis from a binary image is a popular character segmentation method but occasionally fails to segment the characters owing to image noise and uneven illumination. A multimethod binarization scheme that incorporates two or more binary images is a novel solution, but selection of binarization methods has never been analyzed before. This paper reveals the best combination of binarization methods and parameters and presents an in-depth analysis of the multimethod binarization scheme for better character segmentation. We carry out an extensive quantitative evaluation, which shows a significant improvement over conventional single-method binarization methods. Experiment results of six binarization methods and their combinations with different test images are presented.