• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.5
• /
• pp.90-102
• /
• 1993

In this paper, we propose a prgressive image transmission technique using hierarchical pyramid data structure which is constructed based on the projection data of an image. To construct hierarchical Gaussian pyramids, we first divide an image into 4$\times$4 subblocks and generate the projection data of each block along the horizontal, vertical, diagonal, and antidiagonal directions. Among images reconstructed by backprojecting the projection data along a single direction, the one giving the minimum distortion is selected. The Gaussian pyramid is recursively generated by the proposed algorithm and the proposed Gaussian images are shown to preserve edge information well. Also, based on the projection concept a new transmission scheme of the lowest Laplacian plane is presented. Computer simulation shows that the quantitative performance of the proposed pyramid coding technique using projection concept is similar to those of the conventional methods with transmission rate reduced by 0.1 ~ 0.2 bpp and its subjective performance is shown to be better due to the edge preserving property of a projection operation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4439-4455
• /
• 2021

Compared with two-stage object detection algorithms, one-stage algorithms provide a better trade-off between real-time performance and accuracy. However, these methods treat the intermediate features equally, which lacks the flexibility to emphasize meaningful information for classification and location. Besides, they ignore the interaction of contextual information from different scales, which is important for medium and small objects detection. To tackle these problems, we propose an image pyramid network based on dual attention mechanism (DAIPNet), which builds an image pyramid to enrich the spatial information while emphasizing multi-scale informative features based on dual attention mechanisms for one-stage object detection. Our framework utilizes a pre-trained backbone as standard detection network, where the designed image pyramid network (IPN) is used as auxiliary network to provide complementary information. Here, the dual attention mechanism is composed of the adaptive feature fusion module (AFFM) and the progressive attention fusion module (PAFM). AFFM is designed to automatically pay attention to the feature maps with different importance from the backbone and auxiliary network, while PAFM is utilized to adaptively learn the channel attentive information in the context transfer process. Furthermore, in the IPN, we build an image pyramid to extract scale-wise features from downsampled images of different scales, where the features are further fused at different states to enrich scale-wise information and learn more comprehensive feature representations. Experimental results are shown on MS COCO dataset. Our proposed detector with a 300 × 300 input achieves superior performance of 32.6% mAP on the MS COCO test-dev compared with state-of-the-art methods.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.283-294
• /
• 2022

The feature map used in the network for deep learning generally has larger data than the image and a higher compression rate than the image compression rate is required to transmit the feature map. This paper proposes a method for transmitting a pyramid feature map with high compression rate, which is used in a network with an FPN structure that has robustness to object size in deep learning-based image processing. In order to efficiently compress the pyramid feature map, this paper proposes a structure that predicts a pyramid feature map of a level that is not transmitted with pyramid feature map of some levels that transmitted through the proposed prediction network to efficiently compress the pyramid feature map and restores compression damage through the proposed reconstruction network. Suggested mAP, the performance of object detection for the COCO data set 2017 Train images of the proposed method, showed a performance improvement of 31.25% in BD-rate compared to the result of compressing the feature map through VTM12.0 in the rate-precision graph, and compared to the method of performing compression through PCA and DeepCABAC, the BD-rate improved by 57.79%.

• Journal of Biomedical Engineering Research
• /
• v.28 no.2
• /
• pp.199-204
• /
• 2007

In digital radiography, to improve the contrast of digital radiography image, the multi-scale nonlinear amplification algorithm based on unsharp masking is one of the major image enhancement algorithms. In this paper, we used the Laplacian pyramid to decompose a digital radiography(DR) image. In our simulation, the DR image was decomposed into seven layers and the coefficients of the each layer was amplified with nonlinear function. We also imported a noise containment algorithm to limit noise amplification. To enhance the contrast of image, we proposed a new adaptive non-linear gain amplification coefficients. As a result of having applied to some clinical data, a detail visibility was improved significantly without unacceptable noise boosting. Images that acquired with the proposed adaptive non-linear gain coefficients have shown superior quality to those that applied similar gain control method and expected to be accepted in the clinical applications.

• Journal of Sensor Science and Technology
• /
• v.29 no.2
• /
• pp.133-140
• /
• 2020

This study proposes a method of blending visible and near infrared images to enhance edge details and local contrast. The proposed method consists of radiance map generation and color compensation. The radiance map is produced by a Laplacian pyramid and a soft mixing method based on principal component analysis. The color compensation method uses the ratio between the composed radiance map and the luminance channel of a visible image to preserve the visible image chrominance. The proposed method has better edge details compared to a conventional visible and NIR image blending method.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.999-1000
• /
• 2008

Image Registration is used many works in image processing widely. But It is difficult to find the accuracy informations such as translation, rotation, and scaling between images. This paper proposes an algorithm that Fourier based image registration using the pyramid edge detection and line fitting. It can be estimated the informations by each sub-pixels. The proposed algorithm can be used for image registrations which high efficiency is required such as GIS, or MRI, CT, image mosaicing, weather forecasting, etc.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.4
• /
• pp.130-137
• /
• 1993

In this paper, A progressive coding of mean difference pyramid by Hadamard transform of the difference between two successive pyramid levels has been studied. A block adaptive bit allocation method based on ac energy of each sub-block has been proposed, which efficiently reduces the final distortion in the progressive transmission of image parameters. In our scheme, the dc energy equals the sum of the quantization errors of the Hadamard transform coefficients at previous level. Therefore proposed allocation method includes the estimation of dc energy at each pyramid level. Computer simulation results show some improvements in terms of MSE and picture quality over the conventional fixed allocation scheme.

• Proceedings of the KIEE Conference
• /
• 1991.11a
• /
• pp.193-196
• /
• 1991

A method of extracting global, trend curves from input image that may locally not even contain small fragments of those curves using a hierarchical pyramid data structure is suggested. The smoothed input image is subsampled into a pyramid of lower-resolution versions by recursive computation of Gaussian-weighted sums. Trend curves are extracted by finding control points from ridges in these blurred images, and interpolating B-splines for those points.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.3
• /
• pp.268-274
• /
• 2023

Image mosaicking is one of the basic and important technologies in the field of application using images. The key of image mosaicking is to extract seamlines from a joint image. The method proposed in this paper for image mosaicking is as follows. The feature points of the images to be joined are extracted and the joining form between the two images is identified. A reference position for detection the seamlines were selected according to the joint form, and an image pyramid was created for efficient image processing. The outlines of the image including buildings and roads are extracted from the overlapping area with low resolution, and the seamlines are determined by considering the components of the outlines. Based on this, the seamlines in the high-resolution image was re-searched and finally the seamline for image mosaicking was determined. In addition, in order to minimize color distortion of the image with the determined seamline, a method of improving the quality of the mosaic image by fine correction of the mosaic area was applied. It was confirmed that the quality of the seamline extraction results applying the method proposed was reasonable.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 1988.10a
• /
• pp.257-261
• /
• 1988