An Unknown Vehicle Target Recognition Method Based on Radar Broadband Feature
A technology for target recognition and vehicle classification. It is used in character and pattern recognition, instruments, computing, etc., and can solve problems such as increasing computing resource consumption and lengthening algorithm running time.
Active Publication Date: 2022-05-17
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF10 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The process of collecting sample data of unknown categories and training will further lengthen the running time of the algorithm and increase the consumption of computing resources
Method used
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
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0070] Due to the fact that in the actual recognition application, there are unknown types of vehicle targets mixed into the detection target set, resulting in recognition errors, the unknown type of targets should be removed from the target set before identifying the vehicle target. Such as figure 1 As shown, the present invention provides a kind of unknown vehicle target recognition method based on radar broadband feature, and its implementation method is as follows:
[0071] S1. Obtain the radar one-dimensional range image of the known vehicle category and the unknown vehicle category respectively, and use the radar one-dimensional range image of the known vehicle category and the unknown vehicle category as the sample set to be tested, and respectively Label the samples and unknown samples to be tested;
[0072] In this embodiment, the radar one-dimensional range image of the known vehicle category and the unknown vehicle category is used as the sample set to be tested an...
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 PUM
Login to View More Abstract
The invention provides an unknown vehicle target recognition method based on radar broadband features, and belongs to the technical field of unknown vehicle target recognition. First, the radar one-dimensional range images of known vehicle categories and unknown vehicle categories are used as the sample set to be tested; the sample set to be tested is encoded by an autoencoder; the division of known vehicle categories and unknown vehicle categories is calculated using anomaly detection methods Threshold; calculate the discrimination probability of each sample in the sample set to be tested; compare the discrimination probability with the threshold, and eliminate the sample set to be tested according to the comparison result. The present invention is based on the one-dimensional radar high-resolution range image data that has a relatively simple data structure and can be provided all-weather. It uses an autoencoder, error sorting, and anomaly detection algorithms to identify and eliminate unknown vehicle types. Prepare for the training of the class recognition network.
Description
technical field [0001] The invention belongs to the technical field of unknown vehicle target recognition, in particular to an unknown vehicle target recognition method based on radar broadband features. Background technique [0002] The current research on unknown vehicle recognition methods is based on video and pictures as the data sampling equipment and machine learning methods as the technical basis. Taking the recognition method based on video image machine learning as an example, its main idea is to resample and collect the unknown category vehicle data outside the vehicle set to be classified, and input it to the entire generative confrontation network (including generator and Discriminator) to train it, and at the same time jointly use the known category of vehicle video or picture data to train the discriminator network of the generative confrontation network, while improving the network classification ability while preserving the network weight that can discrimina...
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 Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G01S7/41G06V20/54G06V10/774G06V10/764
CPCG01S7/412G06V2201/08G06F18/214
Inventor 冯皓宇张瑛
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA