Visual template matching robustness improving method based on frequency modulation model

A template matching and robust technology, applied in the field of image processing, can solve the problem of low accuracy of image matching, and achieve the effect of improving accuracy and robustness
CN111860643APending Publication Date: 2020-10-30SUZHOU UNIV
3 Cites 4 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SUZHOU UNIV
Publication Date
2020-10-30

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a visual template matching robustness improving method based on a frequency modulation model. The method comprises the following steps: step 1, establishing a topological map by using a RatSLAM algorithm; step 2, in the process of establishing the topological map, performing smoothing processing on the RGB image of the current scene by using a Gaussian function; 3, converting the smoothed image into a Lab color space; 4, calculating an average value of each color channel of the image; 5, calculating the Euclidean distances between the L, a and b values of each pixel andthe L, a and b channel mean values of the image to obtain a saliency map; 6, summing the images according to columns, carrying out normalization processing to obtain a visual template, and storing the visual template in experience points in the topological map; and step 7, matching the visual template by using an SAD (sum of absolute difference) model, and detecting a closed loop so as to correctthe topological map. According to the invention, the robustness of the visual template can be improved, so that the accuracy of image matching is improved.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the field of image processing, in particular to a method for improving the robustness of visual template matching based on a frequency modulation model. Background technique

[0002] Image matching is one of the important methods of closed-loop detection in SLAM (simultaneous localization and mapping), which can reduce the impact of the cumulative error of visual odometry, and is of great significance for mobile robots to establish high-precision maps. In the current RatSLAM algorithm, the closed loop is detected by matching the visual template, thereby reducing the cumulative error caused by the visual odometer. The visual template in the RatSLAM algorithm is obtained by simply summing the grayscale images by column and performing normalization processing, which is easily affected by changes in light intensity, resulting in low accuracy of image matching based on visual templates. Therefore, research on improving the robustne...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More