Abstract
With the rapid development of computer networks and multimedia technologies, images, which are important carriers of information dissemination, have made human cognition of things easier. Image recognition is a basic research task in computer vision, multimedia search, image understanding and other fields. This paper proposes a hierarchical feature learning structure that is completely automatically based on the original pixels of the image, and uses the K-SVD (K-Singular Value Decomposition) algorithm with label consistency constraints to train the discriminant dictionary. For different types of image data sets, the algorithm only extracts image blocks. After dense sampling, an efficient OMP (Orthogonal Matching Pursuit) encoder is used to obtain a layered sparse representation. The improved SIFT (Scale Invariant Feature Transform) algorithm is used to solve the difficult problem of multimedia visual image stereo matching. The feature point extraction and stereo matching of multimedia visual images, different scales and different viewpoint images are analyzed separately. Aiming at a large number of low-dimensional geometric features of 3D images, this paper studies the extraction and sorting strategies of low-dimensional geometric features of 3D images. A sparse representation method for 3D images is proposed, and the sparseness of image features is evaluated. This further improves the accuracy of 3D image representation and the robustness of 3D image recognition algorithms.
Highlights
Image recognition technology has become one of the research hotspots in the field of computer vision in recent years, and has attracted extensive attention from researchers [1]–[3]
In order to solve the problem of huge data volume of low-dimensional geometric features of 3D images, which seriously affects the application of sparse representation, this paper proposes a sorting and selection strategy of 3D image features based on Fisher linear discriminant analysis of sparse representation elements
SIMULATION EXPERIMENT DESIGN AND RESULTS In order to verify the effectiveness of the three-dimensional image representation method, image feature component selection strategy and recognition framework based on sparse representation, a detailed experimental scheme is designed and a large number of comparative experiments are done
Summary
Image recognition technology has become one of the research hotspots in the field of computer vision in recent years, and has attracted extensive attention from researchers [1]–[3]. Relevant scholars have proposed an image recognition method of kernel sparse representation based on dictionary learning [30], [31]. D-KSVD uses linear prediction classification error as the criterion, introduces the discrimination information and classification parameters into the objective function, and uses the K-SVD algorithm to obtain the global optimal solution to all parameters This method cannot guarantee the discriminative power of sparse representation coefficients when using a small dictionary. In order to achieve a balanced reconstruction and discriminativeness, and learn a multi-class linear classifier at the same time, the LC-KSVD method needs to maintain clear consistency between the atoms of the dictionary and class labels This dictionary learning method using supervised information introduces discriminative sparse coding errors and classification errors as regular terms into the objective function. The pixel marked with a cross in the middle layer, if it is the minimum or maximum value of the adjacent 26 pixels DoG, you can use this point as a local extreme point, and record the position of the point and the corresponding scale
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
