An improved method for improving the stability of video object detection

An object detection and stability technology, which is applied in the fields of instruments, character and pattern recognition, computer parts, etc., can solve the problem of insufficient object stability, and achieve the effects of good over-smoothness, speed-up, and wide applicability.

Inactive Publication Date: 2019-06-18
广州二元科技有限公司
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But there is still a problem, that is, the stability of the detected object is not enough. Based on this problem, a better solution is needed to improve the stability of object detection and excessive smoothness in video object detection.

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
  • An improved method for improving the stability of video object detection
  • An improved method for improving the stability of video object detection
  • An improved method for improving the stability of video object detection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0012] The present invention mainly uses the multi-frame object detection results in continuous frame video as a reference, calculates the relationship between the object detection results of the current input frame and the object detection results of the previous frame and multiple frames, and determines the current The output of frame object position.

[0013] There is a common unit of measurement in current display devices: frame rate (FPS), which is the number of frames displayed on the display per second. It is known that the frame rate of ordinary equipment is between 20 and 60, that is to say, the image frames displayed per second are between 20 and 60 frames. This frame rate has great reference value for our subsequent calculations.

[0014] Basically, in the existing software operating systems, there are system-level application programming interfaces (APIs) that can return the current frame rate, and our method also needs to use this frame rate.

[0015] In the pres...

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 relates to an improved method for improving the stability of video object detection. A multi-frame object detection result in a continuous frame video is used as a reference, the relation between a current input frame object detection result and a previous frame and multi-frame object detection result is calculated, and an output result of the current frame object position is determined through calculation of object position offset. According to the improved method for improving the stability of the video object detection provided by the invention, simple mathematical operation is used for carrying out result association calculation between upper and lower frames on the object detection in the video, so that the continuous stability of an object detection frame is ensured; Offsets are calculated on the basis of the weights and the objects storing the FPS number, and better excessive smoothness can be obtained compared with direct averaging; Compared with the prior art, the method has the advantages that complex noise reduction and other processing on an input image are not needed, no special requirement is needed for an object detector, the detection rate can be greatly increased, the applicability is wide, and enough calculation efficiency can be guaranteed in low-end embedded equipment.

Description

technical field [0001] The invention relates to an improved video object detection method for improving stability. Background technique [0002] Moving object recognition and tracking technology in video images is an important topic widely concerned in computer vision, computer science, video surveillance and other academic fields. This technology extends the analysis of static images in the field of digital image processing to dynamic images. The research results of moving object recognition and tracking technology not only have broad application prospects in social daily life and work, but also play an increasingly important role in the fields of military, transportation, medicine and scientific research. The design of moving object recognition and tracking algorithms will directly affect the accuracy and stability of recognition and tracking results. The research on recognition and tracking algorithms with high accuracy and good robustness has always been a hot issue in ...

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): G06K9/00G06K9/62
Inventor 关毅容李庆袁亚荣
Owner 广州二元科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products