Construction method of visual bag-of-words model based on improved surf feature

Abstract

The visual bag-of-words model construction method based on improved SURF features uses the box filter template with gradient information instead of Gaussian filter, which is closer to the Gaussian second-order differential template; when expressing SURF features, it reduces the time overhead, and in While ensuring the invariance of rotation, the SURF descriptor is reduced to 32 dimensions; when constructing the bag of words, the above-mentioned improved SURF algorithm is used to extract all the improved SURF features in the image library, and the k-means clustering method is used to cluster all the SURF features into visual words , so that each image is represented as a high-dimensional vector of the frequency of each visual word. This method contains more abundant gradient information of the image, and omits a Haar wavelet calculation step; compared with directly using SURF features, it can well solve the problem that the number of features extracted from different images is not uniform, and the bag-of-words model can Represent multiple images with a certain amount of visual words, which saves space, is convenient to process, and has strong scalability.

Description

technical field [0001] The invention relates to a method for constructing a visual bag-of-words model based on improved SURF features, and belongs to the technical field of computer vision. Background technique [0002] Compared with the global features of the image, the local features of the image can better describe the image in the face of complex background, large noise interference, changing lighting conditions, multiple things and complex semantics. In recent years, it has been widely used in Image registration, recognition, retrieval, classification and other fields. When directly using local features for image classification and image retrieval, since the number of feature points detected by each image in the image database is not uniform, and commonly used local features such as SIFT, SURF, and DAISY features are high-dimensional features, each image Images are all represented by varying numbers of high-dimensional features, resulting in low efficiency when computi...

AI Technical Summary

Problems solved by technology

The SIFT feature is a highly robust feature extracted based on some local appearance points of interest on the object. It has nothing to do with the size and rotation of the image. It has a high tolerance for light, noise, and micro-angle changes, but the calculation is also relatively complicated. , the efficiency is relatively low

Many scholars have improved the SIFT algorithm. SURF is an improved algorithm of SIFT. The use of integral images and box filters in SURF greatly improves the efficiency of the algorithm, and the processing speed is about three times higher than that of SIFT. However, the SURF algorithm still has certain limitations. Disadvantages: Replacing Gaussian filtering with integral image and box filtering, although the speed is improved, but the gradient information in the image is unavoidably lost; when SURF descriptors are generated, it is necessary to calculate the Haar wavelet response of the circular neighborhood first to obtain the feature points. In the main direction, the Haar wavelet response of the square neighborhood is calculated again to obtain a 64-dimensional vector. This repeated calculation process has a certain impact on the efficiency of SURF

Images