• The Journal of Korean Institute of Information Technology
• /
• v.16 no.12
• /
• pp.85-92
• /
• 2018

When using complicated image processing algorithms, we use superpixels to reduce computational complexity. Superpixel segmentation is a method of grouping pixels having similar characteristics into one group. Since superpixel is used as a preprocessing of image processing, it should be generated quickly, and the edge components of the image should be well preserved. In this paper, we propose a method of generating superpixels with a small amount of computation while preserving edge components well. In the proposed method, superpixels of an image are generated by using the existing k-mean method, and similar superpixels among the generated superpixels are merged to make final superpixels. When merging superpixels, the similarity is calculated only for superpixels. Therefore, the amount of computation is maintained small. It is shown by experimental results that the superpixel images produced by the proposed method are conserving edge information of the original image better than those produced by the existing method.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.06a
• /
• pp.64-65
• /
• 2013

본 논문에서는 슈퍼픽셀(superpixel) 단위의 그래프 컷 알고리즘을 적용하여 객체 추적의 정확도를 향상시키기 위한 방법을 제안한다. 먼저 영상 분할 기법을 사용하여 입력 영상을 슈퍼픽셀로 분할하고 각 슈퍼픽셀에서 색상 히스토그램을 이용한 특성 벡터를 생성한다. 그리고 특성 벡터에 지지벡터기계(support vector machines)를 사용하여 각 슈퍼픽셀의 객체 확률 값을 추정한다. 객체 확률 값을 데이터 항(data term)으로, 이웃한 슈퍼픽셀 간의 특성 벡터 차 값을 스무드 항(smooth term)으로 하여, 그래프 컷(graph cuts) 알고리즘으로 슈퍼픽셀들을 객체와 배경으로 분류하고 객체 슈퍼픽셀을 최대한으로 포함하는 객체 윈도우를 찾는다. 실험 결과는 제안하는 기법이 기존 기법들보다 객체 추적 성능이 우수함을 보여준다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.12
• /
• pp.1520-1527
• /
• 2019

Superpixel methods are widely used in the preprocessing stage as a method to reduce computational complexity by simplifying images while maintaining the characteristics of images in the field of computer vision. It is common to generate superpixels with a regular size and form based on the pixel values rather than considering the characteristics of the image. In this paper, we propose a method to generate superpixels considering the characteristics of an image according to the application. The proposed method consists of two steps, and the first step is to oversegment an image so that the boundary information of the image is well preserved. In the second step, superpixels are merged based on similarity to produce the desired number of superpixels, where the form of superpixels are controlled by limiting the maximum size of superpixels. Experimental results show that the proposed method preserves the boundaries of an image more accurately than the existing method.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.06a
• /
• pp.198-199
• /
• 2016

컴퓨터 비젼 분야에서 이미지 세그멘테이션은 객체 분리, 객체 추적, 의학 영상처리 등 다양한 분야에서 사용된다. 이전의 이미지 세그멘테이션은 사람의 개입이 없이 정확한 객체를 분리하지 못한다는 단점이 있다. 본 논문은 인접한 슈퍼픽셀을 트리를 활용하여 개층적으로 슈퍼픽셀을 통합하는 새로운 세그멘테이션 방법을 소개한다. 제안한 알고리즘을 수행하기 위해 기존의 슈퍼 픽셀 알고리즘을 사용하여, 각 슈퍼픽셀의 센터를 노드로 설정하고 들로네 삼각화를 수행한다. 각각의 인접한 노드는 순차적으로 유사도 측정하여 슈퍼픽셀을 통합한다. 실험 결과를 통해 제안한 방법이 과분할 세그멘테이션을 제거하였으며 영상의 중요한 정보를 잘 보존하는 것을 확인하였다.

• Smart Media Journal
• /
• v.8 no.2
• /
• pp.39-45
• /
• 2019

As deep learning has become popular, researches which can help explaining the prediction results also become important. Superpixel based multi-scale combining technique, which provides the advantage of visual pleasing by maintaining the shape of the object, has been recently proposed. Based on the principle of prediction difference, this technique computes the saliency map from the difference between the predicted result excluding the superpixel and the original predicted result. In this paper, we propose a new technique of both excluding and including super pixels. Experimental results show 3.3% improvement in IoU evaluation.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.10a
• /
• pp.915-917
• /
• 2011

In this paper, a comparison of superpixel characteristics for each color feature space. The superpixel is consist of several pixels with same features such as luminance, color, textures etc. The superpixel can be used on image processing and analysis with large image size to speed up the process. We compare the superpixel characteristics by means of compactness using Berkeley image database(BSD-300).

• Journal of Korea Multimedia Society
• /
• v.15 no.6
• /
• pp.761-769
• /
• 2012

In this paper, a setting method of initial cluster centers and color image segmentation using superpixels and Fuzzy C-means(FCM) algorithm is proposed. Generally, the FCM can be widely used to segment color images, and an element is assigned to any cluster with each membership values in the FCM. However the algorithm has a problem of local convergence by determining the initial cluster centers. So the selection of initial cluster centers is very important, we proposed an effective method to determine the initial cluster centers using superpixels. The superpixels can be obtained by grouping of some pixels having similar characteristics from original image, and it is projected $La^*b^*$ feature space to obtain the initial cluster centers. The proposed method can be speeded up because number of superpixels are extremely smaller than pixels of original image. To evaluate the proposed method, several color images are used for computer simulation, and we know that the proposed method is superior to the conventional algorithm by the experimental results.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.2
• /
• pp.147-157
• /
• 2017

In recent years, Super-pixels have become very popular for use in computer vision applications. Super-pixel algorithm transforms pixels into perceptually feasible regions to reduce stiff features of grid pixel. In particular, super-pixels are useful to depth estimation, skeleton works, body labeling, and feature localization, etc. But, it is not easy to generate a good super-pixel partition for doing these tasks. Especially, super-pixels do not satisfy more meaningful features in view of the gestalt aspects such as non-sum, continuation, closure, perceptual constancy. In this paper, we suggest an advanced algorithm which combines simple linear iterative clustering with fuzzy clustering concepts. Simple linear iterative clustering technique has high adherence to image boundaries, speed, memory efficient than conventional methods. But, it does not suggest good compact and regular property to the super-pixel shapes in context of gestalt aspects. Fuzzy similarity measures provide a reasonable graph in view of bounded size and few neighbors. Thus, more compact and regular pixels are obtained, and can extract locally relevant features. Simulation shows that fuzzy similarity based super-pixel building represents natural features as the manner in which humans decompose images.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.05a
• /
• pp.649-651
• /
• 2012

In this paper, a method of segmenting color image using characteristics of superpixels is proposed. A superpixel is consist of several pixels with same features such as luminance, color, textures etc. The superpixel can be used for image processing and analysis with large scale image to get high speed processing. A color image can be transformed to $La^*b^*$ feature space having good characteristics, and the superpixels are grouped by clustering and gradient-based algorithm.

• Journal of Internet Computing and Services
• /
• v.19 no.3
• /
• pp.43-48
• /
• 2018

In this paper, we propose an indoor-space recognition using DNN and super-pixel. In order to recognize the indoor space from the image, segmentation process is required for dividing an image Super-pixel is performed algorithm which can be divided into appropriate sizes. In order to recognize each segment, features are extracted using a proposed method. Extracted features are learned using DNN, and each segment is recognized using the DNN model. Experimental results show the performance comparison between the proposed method and existing algorithms.