A study of edge preserving filters in image matching

Abstract

This article presents a study on edge preserving filters in image matching which comprises a development of stereo matching algorithm using two edge preserving filters. Fundamentally, the framework is reconstructed by several sequential processes. The output of these processes is a disparity map or depth map. The corresponding points between two images require accurate matching to make accurate depth map estimation. Thus, the propose work in this article utilizes sum of squared differences (SSD) with dual edge preserving filters. These filters are used due to edge preserved properties and to increase the accuracy. The median filter (MF) and bilateral filter (BF) will be utilized. The SSD produces preliminary results with low noise and the edge preserving filters reduce noise on the low texture regions with edge preserving properties. Based on the experimental analysis using the standard benchmarking evaluation system from the Middlebury, the disparity map produced is 6.65% for all error pixels. It shows an accurate edge preserved properties on the disparity maps. To make the proposed work more reliable with current available methods, the quantitative measurement has been made to compare with other existing methods and it displays the proposed work in this article perform much better.

Construction of Automatic Matching Recommendation System for Web Page Image Packaging Design Based on Constrained Clustering Algorithm

“…Hamzah conducted a study on edge preserving filters in image matching. e work proposed by the authors utilizes the sum of squared differences (SSD) and dual edge preserving filters, which effectively preserve the edge properties of the image and improve the matching accuracy [2]. Sadeghi proposed a histogram that combines the advantages of gradient and intensity features (RAGIH).…”

“…(1) Unlike the category label information, the point pair constraint cannot improve the significant data category information in general, it is more generalized in the prior information and belongs to the weaker supervision information. (2) In general, the category information of data points cannot be directly deduced from the point-to-point constraint relationship between the data, which mainly explains whether a pair of data points are connected.…”

“…Because the constrained clustering algorithm has the characteristics of fast matching efficiency, high matching accuracy, and high sensitivity to noise, it improves the performance of the system and makes the design of the system more scientific. (2) e SURF, the SIFT, and the ORB feature description operators are compared and studied to highlight the advantages of the constrained clustering algorithm.…”

Stereo matching algorithm using deep learning and edge-preserving filter for machine vision

“…When deciding on a stereo method, these criteria should be carefully evaluated. The epipolar line is the result of projecting a line from the focus point of the left image onto the image plane of the right image [25]. This line denotes probable equivalent locations in the right image for a pixel in the left image.…”

Refining disparity maps using deep learning and edge-aware smoothing filter

“…Then, calculate the cost function to quantify patch similarity in the input image pair. The second step aggregates costs and filters noise [9]- [11]. Then, winner-take-all (WTA) strategy is employed to select the disparity with the lowest cost while discarding the others.…”