Alternative box updating method and device

An update method and a box-selection technology, applied in the computer field, can solve the problems of increased calculation amount of NMS algorithm, increase of traversal calculation amount, large number of candidate boxes, etc., and achieve the effect of improving work efficiency and efficiency

Active Publication Date: 2019-10-29
HUAWEI TECH CO LTD
View PDF9 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In more complex images, the number of candidate boxes is very large, resulting in an increase in the amount of traversal calculations, which increases the amount of calculations for the NMS algorithm and reduces the work efficiency of image recognition and 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
  • Alternative box updating method and device
  • Alternative box updating method and device
  • Alternative box updating method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] The present application provides a method and device for updating candidate frames, which can be applied to scenes of visual image processing. In the scene of visual image processing, the target object needs to be identified or tracked in the measured image. In order to identify the target object in the measured image, a target recognition frame can be generated. The content in the target recognition frame is the target object in the measured image. In the process of recognizing the target object, multiple recognition frames may be generated for a target object. These multiple recognition frames are all candidates of the target recognition frame, which are referred to as the candidate frame (including the first candidate frame and the first candidate frame). Two alternative boxes). The optimal candidate frame (ie, the first candidate frame) is selected from the candidate frame of a target object as the target candidate frame. In some cases, the tested image may includ...

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 provides an alternative box updating method and device, and relates to the field of computers.The alternative box updating method comprises the steps of obtaining overlapping degree of afirst alternative box and a plurality of second alternative boxes according to position information and position vectors of the first alternative box, wherein the first alternative box is the alternative box with the highest classifier score in the alternative box set, the second alternative box is the alternative box except the first alternative box in the alternative box set, and the position vector comprises position information of the plurality of second alternative boxes; and comparing the overlapping degree of the first alternative box and the plurality of second alternative boxes withthe overlapping degree threshold to obtain an updated alternative box set. By means of the technical scheme, the alternative box updating efficiency can be improved, and then the working efficiency ofimage recognition and detection is improved.

Description

Technical field [0001] This application relates to the computer field, and in particular to a method and device for updating candidate frames. Background technique [0002] Non-Maximum Suppression (English: Non-Maximum Suppression, NMS) is widely used in computer vision algorithms, such as face recognition, edge detection, or target detection. The classifier is an important part of the NMS computer vision algorithm. It can be used to detect an object in an image, such as whether it is a face or not. For an object in the image, the classifier will generate multiple candidate frames. The classifier will calculate the classifier score for each candidate box. In order to be able to accurately detect or recognize objects, only one optimal candidate frame is reserved for an object, and the content in the optimal candidate frame is used as the identified or detected object. [0003] Specifically, select the candidate box with the highest classifier score in all candidate boxes in the im...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62
CPCG06F18/285G06F18/00
Inventor 张锐游骅
Owner HUAWEI TECH CO LTD
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