Method and system for detecting in-vehicle human key points with high efficiency filtering out inter-frame jitter

By using a convolutional network model to determine independent bounding boxes based on the number of seats and performing smoothing filtering in in-vehicle human detection, the problem of high inter-frame jitter complexity in in-vehicle human detection is solved, and efficient and accurate in-vehicle human key point detection is achieved.

CN116434195BActive Publication Date: 2026-08-04MOMENTA (SUZHOU) TECHNOLOGY CO LTD
2 Cites 0 Cited by

Patent Information

Application Number
CN202111657796.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-31
Publication Date
2026-08-04
Estimated Expiration
2041-12-31

AI Technical Summary

Technical Problem

Existing methods for detecting human bodies inside vehicles suffer from high complexity, large computational load, and low accuracy when dealing with inter-frame jitter of human body key points across multiple frames, especially when human bodies overlap or are occluded in the front and rear seats, resulting in a high error rate.

Method used

By acquiring multiple frames of images inside the vehicle using an in-vehicle camera, a convolutional network model is used to perform human body regression processing based on a preset number of seats to determine independent bounding boxes. Smoothing filtering is then applied between the current frame and historical frames to reduce data processing volume and improve detection accuracy.

Benefits of technology

It effectively reduces the difficulty of model processing, reduces the amount of data processing, improves the accuracy of human key point detection, and efficiently eliminates inter-frame jitter.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application discloses a kind of high-efficiency filtering interframe jitter's in-car human body key point detection method and system, belong to computer vision technical field.The method includes: by vehicle-mounted camera is photographed in-car, obtains in-car multiple frames of pictures;Using pre-trained convolution network model, the current frame of picture is handled according to preset seat number Human regression, determine the preset seat number personal body bounding box on current frame of picture;The human body in human body bounding box is detected, determine human body key point;The human body key point in each human body bounding box of current frame of picture is carried out with the human body key point in the corresponding human body bounding box of historical frame of picture Smooth filtering processing.The application determines a certain number of human body bounding boxes according to the number of in-car seats, reduces the difficulty of model processing problem, reduces data processing amount, simultaneously using a certain number, independent human body bounding box is associated between multiple frames The human body key point belonging to the same human body is efficiently eliminated interframe jitter.
Need to check novelty before this filing date? Find Prior Art