Video frame optimization method and device

A technology of video frame and target video, which is applied in the field of cameras, can solve problems such as complex implementation, achieve the effect of improving performance and realizing automatic learning

Pending Publication Date: 2021-06-04
HANGZHOU HIKVISION DIGITAL TECH
View PDF6 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, in practice, it has been found that in the current optimal realization schemes of goals, it is necessary to manually participate in the design of complex comparison logic, or to design a combination of comprehensive scores, which is relatively complicated to achieve

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
  • Video frame optimization method and device
  • Video frame optimization method and device
  • Video frame optimization method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0105] Embodiment 1. Each human face video frame sequence retains one frame of the best image

[0106] See Figure 3A , is a closed-loop schematic diagram of reinforcement learning in this embodiment, such as Figure 3A As shown, in this embodiment, the state quantities of the agent include the score combination of the optimal frame and the score combination of the currently input face video frame; each: renewal and abandonment; of which:

[0107] The update indicates that the agent believes that the currently input face video frame is more conducive to recognition, thereby replacing the existing optimal frame with the currently input face video frame (synchronously changing the optimal frame state in the state);

[0108] Abandoning indicates that the agent believes that the current input face video frame is not as good as the existing optimal frame for recognition.

[0109] In this embodiment, the similarity between each face video frame and the true value of the face is u...

Embodiment 2

[0111] Embodiment two, each face video frame sequence retains multiple frames (not more than the preset upper limit) images

[0112] Please refer to 3B, which is a closed-loop schematic diagram of reinforcement learning in this embodiment, such as Figure 3B As shown, in this embodiment, the state quantity of the agent includes the number of reserved frames, the score combination of each reserved frame and the score combination of the currently input face video frame; the involved response actions include four: add, update , abandonment and deletion; where:

[0113] The new addition indicates that the number of reserved frames has not reached the upper limit, and the agent believes that adding the currently input face video frame to the reserved frame is more conducive to recognition, so it adds the currently input face video frame to the reserved frame (synchronously update the number of reserved frames, each reserved frame frame's scoring combination);

[0114] The update ...

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 a video frame optimization method and device, and the method comprises: executing a preset number of times of optimal frame selection actions based on a currently used optimal frame selection strategy in an optimal frame selection strategy training process, and updating the currently used optimal frame selection strategy based on a return value corresponding to the preset number of times of optimal frame selection actions until a preset strategy updating completion condition is reached; and performing video frame optimization on the target video frame sequence to be optimized based on a trained optimal frame selection strategy and the state quantity of each target video frame in the target video frame sequence to be optimized. According to the method, the difficulty in design of a combination mode of comparison logic or comprehensive scoring can be avoided, automatic learning of face optimization is realized, and the face optimization performance is improved.

Description

technical field [0001] The present application relates to the technical field of cameras, and in particular to a video frame optimization method and device. Background technique [0002] Target selection is a traditional concept in target recognition. Its purpose is to optimize the quality of the target video frame sequence obtained by tracking, and to select video frames suitable for target recognition (usually only one frame or a small number of frames are selected for a sequence. ), so as to effectively reduce the amount of data to be processed in subsequent recognition steps, and also ensure the performance of the overall recognition system. [0003] Taking face selection as an example, there are many factors that affect face recognition, usually including at least the size of the face, the angle of the face, the clarity of the face, the occlusion of the face, and the lighting conditions. Quantitative evaluation of face recognition factors is a prerequisite for face sel...

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): G06K9/00
CPCG06V40/161G06V20/41
Inventor 王春茂浦世亮潘之玮
Owner HANGZHOU HIKVISION DIGITAL TECH
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap