Image data set redundancy cleaning method for safety helmet wearing detection
A technology of image data sets and safety helmets, which is applied to instruments, character and pattern recognition, computer components, etc., and can solve problems such as higher structural similarity values
Pending Publication Date: 2022-02-18
东南数字经济发展研究院 +1
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, if it is applied to the helmet wearing monitoring project, the structural similarity value will become higher due to the change of the non-hard hat area, and it will be mistakenly judged as a difference between the two pictures. This is often what we do not want to see of
[0010] The above method does not work well for redundant cleaning of image datasets in helmet wearing detection projects
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0035] The present invention will be described in further detail below in conjunction with the accompanying drawings.
[0036] combined with figure 1 ,
[0037] Implementation column one
[0038] A method for redundant cleaning of image datasets for helmet wearing detection, comprising the following steps:
[0039] S1, use the target detection algorithm YOLO V4 algorithm to train on the VOC data set, use the back propagation algorithm to optimize the parameters, and select the model file with the best result to save;
[0040] S2, using FFmpeg to parse the video captured by the camera into an image;
[0041] S3, the analyzed images are input in order to the target detection model trained in step S1, and the detection results only retain the images in which people are detected;
[0042] S4, judging whether the number of people detected in the two pictures before and after is consistent;
[0043] S5, if the judgment result in step S4 is inconsistent, then both images are sav...
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
Login to View More Abstract
The invention relates to an image data set redundancy cleaning method for safety helmet wearing detection. The method comprises the following steps of: carrying out training on a VOC data set through employing a target detection algorithm YOLO V4, and selecting a model file with an optimal result for storage; analyzing a video acquired by a camera into images by using FFmpeg; inputting the analyzed images into a trained target detection model in sequence, and enabling the detection result to only retain images with persons; judging whether the number of persons detected in the two consecutive images is consistent; if not, storing the two images; if yes, calculating the coordinates of the position center point of each person in the two images; calcualting the sum of the distances between the corresponding center points; judging whether the sum of the calculated distances is greater than a set threshold; if not, deleting one of the pictures; and if yes, retaining the two images. The method has the advantage that the interpretability of the safety helmet wearing detection image data redundancy cleaning method is improved.
Description
technical field [0001] The invention relates to the technical field of target detection, in particular to a method for redundant cleaning of image data sets for helmet wearing detection. Background technique [0002] As a work helmet that must be worn during construction operations, safety helmets have many applications, such as construction sites, factory production lines, and underground mining. Wearing safety helmets by operators is not only a standard requirement for safe production operations, but also an important guarantee for personal safety. Therefore, the helmet wearing detection model has a very broad application prospect. At the same time, the accuracy of the helmet wearing detection model needs to reach a certain level. This requires targeted helmet wearing detection model training for the surveillance video data of each specific application. [0003] With the continuous development of deep learning technology, the target detection model is one of the importa...
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
Login to View More IPC IPC(8): G06V10/72G06V10/74G06V10/774G06V20/52G06V20/40G06K9/62
CPCG06F18/22G06F18/10G06F18/214
Inventor 陈轶张文王茂森牛少彰崔浩亮冯亚辉王让定
Owner 东南数字经济发展研究院