scispace - formally typeset
Search or ask a question
Topic

Bilateral filter

About: Bilateral filter is a research topic. Over the lifetime, 3500 publications have been published within this topic receiving 75582 citations.


Papers
More filters
Proceedings ArticleDOI
19 Dec 2015
TL;DR: The experiment result shows that the proposed algorithm achieves a satisfied image enhancement effect as well as eliminates the halation, and its processing speed is obviously faster than both the bilateral algorithm and the traditional Retinex algorithm.
Abstract: Retinex is a widely used method in image enhancement, but it may cause halation in the high contrast area. Bilateral filter is an effective algorithm to eliminate this phenomenon, but its slow processing speed restricts its application. To solve this problem, a new Retinex algorithm based on recursive bilateral filter is proposed in this paper. Firstly, recursive bilateral filter is used to estimate the illumination image, and then we use a traditional Retinex module to get the reflectance image and perform a histogram equalization to improve its visual effect. At last, this algorithm is evaluated from both subject and objective aspects. The experiment result shows that the proposed algorithm achieves a satisfied image enhancement effect as well as eliminates the halation, and its processing speed is obviously faster than both the bilateral algorithm and the traditional Retinex algorithm.

12 citations

Journal ArticleDOI
TL;DR: The fast bilateral filter is employed for noise removal and it has good edge preservation capacity and the canny edge detector is efficient when compared with the conventional edge detectors.
Abstract: The role of preprocessing and segmentation are vital in image processing and computer vision. The medical images are prone to noise and the filtering algorithms are used for noise removal. In this paper, the fast bilateral filter is employed for noise removal and it has good edge preservation capacity. The segmentation algorithms are used to extract the region of interest and edge detection is a classical algorithm for tracing the contours of objects in an image. The canny edge detector is efficient when compared with the conventional edge detectors. The fast bilateral filter is proposed in this paper has the computation complexity of O(1) per pixel, while the classical bilateral filter has the computation complexity of O(W) operations per pixel, where W is the kernel size. The algorithms were implemented in Raspberry Pi using Open CV software package. The algorithms were tested on real time medical images.

12 citations

Proceedings ArticleDOI
01 Oct 2006
TL;DR: A new fuzzy filter for a more general noise model in which a noisy pixel has an arbitrary value in the dynamic range according to some underlying probability distribution is presented.
Abstract: In many current impulse noise models for images, corrupted pixels are replaced with values equal or near the maximum or minimum intensity values of the allowable dynamic range. In this paper, we present a new fuzzy filter for a more general noise model in which a noisy pixel has an arbitrary value in the dynamic range according to some underlying probability distribution. This filter consists of (1) a fuzzy detection method, where we investigate if a certain pixel position can be seen as noisy or not and (2) a fuzzy reduction method that reduces the noise while preserving the fine details (like edges and textures) of the image. Experimental results have shown that the proposed filter may be used for efficient removal of randomly valued impulse noise without distorting the useful information in the image.

12 citations

Proceedings ArticleDOI
03 Jun 2012
TL;DR: The experimental results indicate that the proposed algorithm is able to provide dense disparity maps with sub-pixel resolution and achieves better accuracy compared to two similar stereo matching algorithms.
Abstract: This paper presents a local disparity calculation algorithm on calibrated stereo images based on cost aggregation. Unlike most of the existing cost aggregation methods which are mainly based on the grouping of colour similarities, the proposed algorithm is grouped by local cost similarities. The proposed algorithm also applies a bilateral filter to enhance the normalised cost volume and, then, uses the winner-take-all technique to select the correspondence candidates. Finally, a quadratic polynomial interpolation is performed using the candidates and their neighbourhood values to achieve sub-pixel disparity resolution. The experimental results indicate that the proposed algorithm is able to provide dense disparity maps with sub-pixel resolution and achieves better accuracy compared to two similar stereo matching algorithms.

12 citations

Book ChapterDOI
04 Oct 2020
TL;DR: JBFnet as discussed by the authors is a neural network for low-dose CT denoising, where the guidance image is estimated by a deep neural network and the filter functions of the joint bilateral filter are learned via shallow convolutional networks.
Abstract: Deep neural networks have shown great success in low dose CT denoising. However, most of these deep neural networks have several hundred thousand trainable parameters. This, combined with the inherent non-linearity of the neural network, makes the deep neural network difficult to understand with low accountability. In this study we introduce JBFnet, a neural network for low dose CT denoising. The architecture of JBFnet implements iterative bilateral filtering. The filter functions of the Joint Bilateral Filter (JBF) are learned via shallow convolutional networks. The guidance image is estimated by a deep neural network. JBFnet is split into four filtering blocks, each of which performs Joint Bilateral Filtering. Each JBF block consists of 112 trainable parameters, making the noise removal process comprehendable. The Noise Map (NM) is added after filtering to preserve high level features. We train JBFnet with the data from the body scans of 10 patients, and test it on the AAPM low dose CT Grand Challenge dataset. We compare JBFnet with state-of-the-art deep learning networks. JBFnet outperforms CPCE3D, GAN and deep GFnet on the test dataset in terms of noise removal while preserving structures. We conduct several ablation studies to test the performance of our network architecture and training method. Our current setup achieves the best performance, while still maintaining behavioural accountability.

12 citations


Network Information
Related Topics (5)
Image processing
229.9K papers, 3.5M citations
87% related
Image segmentation
79.6K papers, 1.8M citations
87% related
Feature (computer vision)
128.2K papers, 1.7M citations
86% related
Feature extraction
111.8K papers, 2.1M citations
86% related
Pixel
136.5K papers, 1.5M citations
84% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
202321
202257
2021116
2020145
2019203
2018204