• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2006년도 Proceedings of ISRS 2006 PORSEC Volume II
• /
• pp.736-739
• /
• 2006

While hyperspectral data are very rich in information, their processing poses several challenges such as computational requirements, noise removal and relevant information extraction. In this paper, the application of advanced scale-space filtering to selected hyperspectral bands was investigated. In particular, a pre-processing tool, consisting of anisotropic diffusion and morphological leveling filtering, has been developed, aiming to an edge-preserving smoothing and simplification of hyperspectral data, procedures which are of fundamental importance during feature extraction and object detection. Two scale space parameters define the extent of image smoothing (anisotropic diffusion iterations) and image simplification (scale of morphological levelings). Experimental results demonstrated the effectiveness of the developed scale space filtering for the enhancement and smoothing of hyperspectral remote sensing data and their advantage against watershed over-segmentation problems and edge detection.

• 한국전자통신학회논문지
• /
• 제14권2호
• /
• pp.433-438
• /
• 2019

본 논문에서는 기계 서비스 부품 도면에서 숫자를 인식하기 위하여 입력 영상에 대한 전처리와 딥러닝 모델을 제안한다. 서비스 부품 도면의 숫자를 인식하는데 있는 지시선과 도형에 의한 오검출 또는 오인식을 개선하기 위하여 수학적 형태학 필터링 전처리를 한다. 숫자 인식을 위하여 VGG-16 모델을 축소 변형한 7 개의 계층을 가지는 VGG 모델을 적용함으로써 인식 성능을 개선한다. 서비스 부품 도면의 숫자 인식 실험 결과, 제안하는 방법이 인식률 95.57%, 정확도는 92.82%로 종래의 방법에 현저히 개선된 결과를 얻었다.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 1998년도 Proceedings of International Workshop on Advanced Image Technology
• /
• pp.29-33
• /
• 1998

In this paper, we propose a hierarchical segmentation method for tracking a semantic video object using a watershed algorithm based on morphological filtering. In the proposed method, each hierarchy consists of three steps: First, markers are extracted on the simplified current frame. Second, region growing by a modified watershed algorithm is performed for segmentation. Finally, the segmented regions are classified into 3 categories, i.e., inside, outside, and uncertain regions according to region probability values, which are acquired by the probability map calculated from a estimated motion field. Then, for the remaining uncertain regions, the above three steps are repeated at lower hierarchies with less simplified frames until every region is decided to a certain region. The proposed algorithm provides prospective results in video sequences such as Miss America, Clair, and Akiyo.

• 전자공학회논문지B
• /
• 제33B권7호
• /
• pp.116-126
• /
• 1996

In this paper, we propose a new termination rule for the iterative restoration of degraded images, by which the number of iterations can be dramatically reduced. This rule uses a parameter, called noise suppression factor (NSF), to appropriately teminate the iteration process, reduces a lot the computational load, and avoids the amplification of noise for the better quality. We also propose a method using the morphological filters, when applied to the resulting image, that will significantly reduce the ringing effect which would otherwise exist in the boundary of the image. Simulation with the blurred lena image with gaussian noise showns that the proposed termination rule combined with the morphological filtering gives the restored image with much improved quality.

• Journal of Information Processing Systems
• /
• 제16권6호
• /
• pp.1424-1436
• /
• 2020

Although huge progress has been made in current image segmentation work, there are still no efficient segmentation strategies for tree image which is taken from natural environment and contains complex background. To improve those problems, we propose a method for tree image segmentation combining adaptive mean shifting with image abstraction. Our approach perform better than others because it focuses mainly on the background of image and characteristics of the tree itself. First, we abstract the original tree image using bilateral filtering and image pyramid from multiple perspectives, which can reduce the influence of the background and tree canopy gaps on clustering. Spatial location and gray scale features are obtained by step detection and the insertion rule method, respectively. Bandwidths calculated by spatial location and gray scale features are then used to determine the size of the Gaussian kernel function and in the mean shift clustering. Furthermore, the flood fill method is employed to fill the results of clustering and highlight the region of interest. To prove the effectiveness of tree image abstractions on image clustering, we compared different abstraction levels and achieved the optimal clustering results. For our algorithm, the average segmentation accuracy (SA), over-segmentation rate (OR), and under-segmentation rate (UR) of the crown are 91.21%, 3.54%, and 9.85%, respectively. The average values of the trunk are 92.78%, 8.16%, and 7.93%, respectively. Comparing the results of our method experimentally with other popular tree image segmentation methods, our segmentation method get rid of human interaction and shows higher SA. Meanwhile, this work shows a promising application prospect on visual reconstruction and factors measurement of tree.

• 대한방사선기술학회지:방사선기술과학
• /
• 제45권5호
• /
• pp.397-405
• /
• 2022

Computed radiography (CR) systems, which convert an analog signal recorded on a cassette into a digital image, combine the characteristics of analog and digital imaging systems. Compared to digital radiography (DR) systems, CR systems have presented difficulties in evaluating system performance because of their lower detective quantum efficiency, their lower signal-to-noise ratio (SNR), and lower modulation transfer function (MTF). During the step of energy-storing and reading out, a baseline offset occurs in the edge area and makes low-frequency overestimation. The low-frequency offset component in the line spread function (LSF) critically affects the MTF and other image-analysis or qualification processes. In this study, we developed the method of baseline correction using mathematical morphology to determine the LSF and MTF of CR systems accurately. We presented a baseline correction that used a morphological filter to effectively remove the low-frequency offset from the LSF. We also tried an MTF evaluation of the CR system to demonstrate the effectiveness of the baseline correction. The MTF with a 3-pixel structuring element (SE) fluctuated since it overestimated the low-frequency component. This overestimation led the algorithm to over-compensate in the low-frequency region so that high-frequency components appeared relatively strong. The MTFs with between 11- and 15-pixel SEs showed little variation. Compared to spatial or frequency filtering that eliminated baseline effects in the edge spread function, our algorithm performed better at precisely locating the edge position and the averaged LSF was narrower.