Identification method for dressing specifications of distribution network field operating personnel
A technology of on-site operation and recognition method, which is applied in biometric recognition, neural learning method, character and pattern recognition, etc., and can solve problems such as poor real-time performance, incomplete coverage, and heavy workload
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0037] The technical scheme of the present invention comprises the following steps:
[0038] Image classification: intercept the video of the field operation of the distribution network as an image set, and divide the image set into a verification set and a training set according to the proportion;
[0039] In the image classification step, the processing and labeling of the image samples are also included, and the processing is to scale the set of images to a fixed size.
[0040] Specifically, the image samples are scaled by using the bilinear interpolation algorithm, so that the image set is scaled to a fixed size.
[0041] Further, the bilinear interpolation algorithm is specifically, the f function P=(x, y) is the interpolation coordinate point to be solved, and it is known that point P is the four coordinate points Q around the field shape. 11 =(x 1 , y 1 ,),
[0042] Q 12 =(x 1 y 2 ), Q 21 =(x 2 , y 1 ), Q 22 =(x 1，2 y 2 ), using bilinear interpolation as fo...
Embodiment 2
[0062] In this embodiment, the description is made with reference to the specific practical use.
[0063] Refer to the attached Figure 1-5 As shown, according to the complete method embodiment of the present invention and its implementation process as follows:
[0064] 1) The maintenance work video of the distribution network field operators is collected through the work recorder, and uploaded to the monitoring center, and the video of the monitoring center is cut into pictures by frame. The captured images are marked as not wearing work clothes and wearing work clothes, and there can be multiple targets and two categories in one image. 12,000 pictures were intercepted as the training data set, 6,000 pictures with work clothes and 6,000 pictures without work clothes.
[0065] 2) Scale the marked image, and use the bilinear interpolation algorithm to scale the image to a fixed sample data set.
[0066] 3) Use the labeling tool to label, label whether the workers in the samp...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More