Color texture feature extraction method based on quaternion gabor filter
A texture feature and extraction method technology, applied in the field of image processing, can solve the problems of not using texture image color features, not providing color texture feature images, ignoring image color features, etc.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Publication Date
- 2021-09-28
Smart Images

Figure 1 
Figure 2 
Figure 3
Abstract
Description
technical field
[0001] The invention belongs to the technical field of image processing, and in particular relates to a color texture feature extraction method based on quaternion Gabor filtering. Background technique
[0002] Texture is an important visual clue of features, and an important but difficult to describe feature in image processing. Up to now, there is still no recognized and strict definition of image texture features. The research content of texture analysis mainly includes texture classification and segmentation, texture synthesis, texture retrieval and texture restoration shape. One of the most basic problems in these research contents is texture feature extraction. The quality of extracted texture features directly affects the segmentation and classification results. The existing texture feature extraction methods mainly include statistical methods, model methods, signal processing methods and structural methods. Among them, the Gabor filter in the signal ...
Examples
Embodiment 1
[0041] Color texture feature extraction method based on quaternion Gabor filter, described color texture feature extraction method is to utilize quaternion Gabor filter and quaternion Gabor convolution algorithm to carry out feature extraction to color image, obtain 5 according to traditional Gabor filter setting 40 color texture feature images of 40 scales and 8 directions (the scales and directions here are not limited to specific 5 scales and 8 directions, in other embodiments, other values can also be taken); then, calculate For each feature image and the Tamura texture features (roughness, contrast, directionality) of the original image, according to the Euclidean distance, select the 3 feature images with the highest similarity to the original image (requiring different scales and different directions), This imitates image rotation and zooming in and out, and finally calculates the Tamura texture features of the three feature images and the color components of the color...