Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Rapid vehicle detection method

A vehicle detection and detection method technology, applied in the direction of instruments, character and pattern recognition, computer components, etc., can solve the problems of vehicle detection, mutual occlusion vehicles are difficult to accurately detect, etc., to prevent the gradient from disappearing, and the characteristics of the window are simple , the effect of less dependence on the environment

Pending Publication Date: 2019-05-31
QINGDAO UNIV
View PDF0 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art, seek to design a fast vehicle detection method, based on the window feature as the detected object, combined with the residual module of ResNet (residual connection network) and SSD (SingleShot Detector) algorithm The multi-scale feature extraction process, drawing on the network structure of YOLOv3, constructs a full convolution detection method with only 24 convolutional layers, which can better solve the problem that it is difficult to accurately detect mutual occlusion vehicles in the detection of moving vehicles on the road. The detection algorithm of deep learning cannot detect the vehicles in the surveillance video in real time

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Rapid vehicle detection method
  • Rapid vehicle detection method
  • Rapid vehicle detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] This embodiment relates to a rapid vehicle detection method and its application test, and its specific process includes the following steps:

[0034] (1) Data collection: due to various complex traffic conditions, weather conditions, and different time periods, it will have different effects on the number of target vehicles, distribution status, brightness, and color saturation of the surveillance video; in order to make the detection method applicable For various road scenarios, when collecting training data, a large number of pictures of different road sections, different road conditions, different weather, and different time conditions are selected from the traffic monitoring video. The traffic flow in the pictures is relatively large, and the vehicles contained in each picture The number is more than 25, and the size of the picture is 1920px×1080px, which is used as data;

[0035] (2) Data labeling: do data labeling on the pictures collected in step (1), and use the...

Embodiment 2

[0048] This embodiment selects the video screenshot taken by the video monitoring equipment at the higher position of the two-way lane, compares the detection method of the present invention with the Faster R-CNN detection algorithm, see figure 2 ;in, figure 2 .(a) It can be seen that there are three cars parked on the roadside on the left side of the road, which are blocked by tree branches. There are many vehicles on the road, and there are mutual occlusions; figure 2 .(b) It can be seen that the recognition effect of the Faster R-CNN detection algorithm is not ideal, the detection rate is 55%, and the detection accuracy is 100% detection accuracy. Mutual occlusion causes two vehicles to be identified as one vehicle, which will have a great impact on the statistics of traffic flow, and its detection speed is slower than that of this detection method; figure 2 .(c) It can be seen that the detection method of the present invention has better performance, and the vehicles ...

Embodiment 3

[0050] In this embodiment, the video surveillance shooting at a relatively short distance is selected, and the traffic flow is relatively large, and the vehicles in the area in the camera block each other seriously, such as image 3 (a); Contrast the detection method of the present invention with the Faster R-CNN detection algorithm, image 3 .(b) It can be seen that Faster R-CNN has a detection rate of 57% for vehicles with serious occlusion problems, and a detection accuracy of 85%. It is common to recognize multiple vehicles as one vehicle, and its IOU (Intersection over Union) has a large error; image 3 .(c) It can be seen that the detection effect of this method is better than that of the existing technology, the detection rate is 79%, and the detection accuracy is 100%. Vehicles on the boundary are not detected, and other vehicles that block each other are very clear Clearly detect and label.

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

No PUM Login to View More

Abstract

The invention belongs to the technical field of video detection in deep learning. The invention relates to a vehicle detection method, in particular to a rapid vehicle detection method based on windowcharacteristics. According to the method, a vehicle window is used for replacing a vehicle body to serve as a target object for detection, a residual module of a ResNet network and a multi-scale feature extraction method of an SSD algorithm are combined, a network structure of YOLOv3 is used for reference, and a full convolution detection method with only 24 convolution layers is constructed; Under the condition that the traffic flow is large, during batch testing, the average detection precision is close to 100%, the average detection rate reaches 90%, the detection speed reaches 22 milliseconds per frame, real-time detection of vehicles in a road high-definition monitoring video is achieved, the detection rate of the vehicles in the large traffic flow is effectively increased, and the method has important application value.

Description

Technical field: [0001] The invention belongs to the technical field of video detection in deep learning, and in particular relates to a fast vehicle detection method based on vehicle window features, in particular to a method capable of quickly realizing the state detection of a moving vehicle. Background technique: [0002] At present, deep learning algorithms based on Convolutional Neural Networks (CNN) are developing rapidly, and relying on their powerful feature extraction capabilities, they have achieved high accuracy in target detection and recognition. However, the current detection algorithm based on deep learning has a major disadvantage, that is, the problem of speed, such as Faster R-CNN and other algorithms with region proposal (selecting the region of interest in advance, and then further detection) step, the detection accuracy is very high, such as the Chinese patent CN2018106089322 discloses a large-scale construction vehicle boom detection algorithm, which c...

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
no application Login to View More
IPC IPC(8): G06K9/00
CPCY02T10/40
Inventor 王国栋王亮亮潘振宽徐洁王岩杰李宁孝胡诗语
Owner QINGDAO UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com