Multi-view model recognition method based on integrated local linear embedding and linear discrimination analysis

Abstract

The present invention provides an En-ULLELDA-based (ensemble-unified locally linear embedding and linear discriminant analysis) method of multi-view model recognition. The method comprises the following steps: firstly, selecting a vehicle ROI (region of interest); secondly, extracting the high-level information that can reflect the model as features through Gabor wavelet transform; then, carrying out the dimension reduction of features on the basis of a novel manifold learning algorithm En-ULLELDA, so as to transform the features to a low-dimensional space; and obtaining the corresponding models from the features of a test sample by a classifier during the recognition. According to test results, the invention can effectively handle the problem of multi-view model recognition on urban roads; and the invention can eliminate the effects of parameter variation on the recognition results by employing the integration technology, so that the invention has higher robustness.

Description

technical field [0001] The invention belongs to an image recognition method, and in particular relates to a multi-view vehicle vehicle recognition method based on the combination of integrated local linear embedding and linear discriminant analysis (En-ULLELDA). Background technique [0002] Vehicle model recognition refers to the identification of the manufacturer and model of the vehicle through the collected vehicle images or information. For example, Audi (manufacturer) A6 (model). Vehicle type recognition has a strong application background, such as investigating the market share of a certain type of vehicle, finding vehicles of a certain type passing through intersections or toll booths to assist the police in solving cases, etc. Vehicle type recognition is a pattern recognition problem. At present, there are relatively few studies on vehicle type recognition, and only a few researchers have done research in this area. The current vehicle identification methods can b...

AI Technical Summary

Problems solved by technology

However, these three methods have limitations in recognition rate and practical application, and are suitable for vehicle recognition under fixed viewing angles such as parking lots and toll stations.

[0003] In practical applications, such as urban roads, highways, etc., the camera set up above the road has a wide field of view, so the angle of view of the vehicle entering the camera's field of view will be different due to different positions and different vehicle directions.

In this way, only extracting ROI based on the located license plate will produce differences, which will affect the recognition results

That is to say, the vehicle identification method with a fixed perspective is no longer suitable for solving the problem of multi-view. Therefore, it is necessary to study the vehicle identification technology in the case of multi-view. There are even fewer methods to identify

Images