Abstract:Image smoothing is a fundamental task in signal processing. For such task, box filter is well-known. However, box filter can not keep some features of the signal, such as edges, corners and the jump in the step function. In this paper, we present a one-sided box filter that can smooth the signal but keep the discontinuous features in the signal. More specifically, we perform box filter on eight one-sided windows, leading to a one-sided box filter that can preserve corners and edges. Our filter inherits the constant $O(1)$ computational complexity of the original box filter with respect to the window size and also the linear $O(N)$ computational complexity with respect to the total number of samples. We performance several experiments to show the efficiency and effectiveness of this filter. We further compare our filter with other the-state-of-the-art edge preserving methods. Our filter can be deployed in a large range of applications where the classical box filter is adopted.

What problem does this paper attempt to address?

### The Problem the Paper Attempts to Solve This paper aims to address a key issue in image smoothing: how to preserve edges and corner features in an image while smoothing it. Traditional Box Filters perform well in image smoothing but fail to retain these important features, such as edges, corners, and discontinuities in step functions. To solve this problem, the authors propose a new filtering method—the One-Sided Box Filter, which can effectively preserve these features while smoothing the image. ### Specific Problem Description 1. **Importance of Image Smoothing**: - Image smoothing is a fundamental task in signal processing, widely used in image denoising, compression, interpolation, segmentation, deconvolution, enhancement, and other tasks. - Smoothing can remove noise and unnecessary details from an image, which is crucial for accomplishing the aforementioned tasks. 2. **Limitations of Traditional Box Filters**: - Box Filters are a commonly used smoothing method, but they cannot preserve important features in an image, such as edges and corners. - This is because Box Filters are an isotropic diffusion process that smooths out all features in the image. 3. **Proposed New Method**: - The authors propose a One-Sided Box Filter, which performs box filtering on eight one-sided windows to achieve image smoothing. - This method inherits the constant complexity $O(1)$ in window size and linear complexity $O(N)$ in the total number of samples from traditional Box Filters. - The One-Sided Box Filter can preserve edges and corner features in the image, making it suitable for applications that require classic Box Filters. ### Characteristics of the Solution - **Computational Efficiency**: The One-Sided Box Filter maintains computational efficiency while effectively preserving important features in the image. - **Adaptability**: This method is suitable for various image processing tasks, including image denoising, enhancement, and segmentation. - **Theoretical Guarantee**: The edge and corner preservation characteristics of the One-Sided Box Filter are theoretically guaranteed, without the need for manual parameter adjustment. ### Experimental Validation - The authors demonstrate the effectiveness and efficiency of the One-Sided Box Filter through multiple experiments and compare it with other state-of-the-art edge-preserving methods. - Experimental results show that the One-Sided Box Filter excels in preserving image features and has high computational efficiency. ### Conclusion This paper proposes a new One-Sided Box Filter that addresses the issue of traditional Box Filters failing to preserve edges and corner features during image smoothing. This method not only matches the computational efficiency of traditional Box Filters but also excels in preserving image features, showing great potential for a wide range of applications.

One-Sided Box Filter for Edge Preserving Image Smoothing

L0 image smoothing via iterating truncated L1 gradient regularization

Edge-Preserving Piecewise Linear Image Smoothing Using Piecewise Constant Filters

Fast algorithm for edge-preserving multi-scale image decomposition

Quarter Laplacian Filter For Edge Aware Image Processing

Modified Edge-Oriented Spatial Interpolation For Consecutive Blocks Error Concealment

A Parameter Adaptive Method for Image Smoothing

A Robust Blind Deblurring Method for Motion Images Based on Edge Information

Embedding Bilateral Filter in Least Squares for Efficient Edge-Preserving Image Smoothing

Image Edge Restoring Filter

Smoothing Algorithm with Edge-preserving by Extrema Constraints

Edge-Aware Volume Smoothing Using L0 Gradient Minimization

Real-Time Image Smoothing Based on True Edges

Edge-Respecting Image Smoothing via Extrema Interpolation.

Adaptive Image Smoothing Based on Robust Estimation

Edge-preserving Smoothing for Image Decomposition Via a Hybrid Approach

Feature-preserving filtering with L0 gradient minimization

Edge-preserving Image Filtering Based on Soft Clustering

Real-time Image Smoothing via Iterative Least Squares

Enhanced Edge-Perceptual Guided Image Filtering

Two-Pass Bilateral Smooth Filtering for Remote Sensing Imagery