Automatic labeling method of visibility grades in highway fog images

Abstract

The present invention provides a method for automatically labeling the visibility level of expressway group fog pictures, including the following steps: step S1, using the marked group fog picture test sample set to test a plurality of known group fog detection algorithms, and obtain multiple The recognition accuracy matrix A of the known group fog detection algorithm; step S2, using the multiple known group fog detection algorithms to identify the group fog picture to be marked to obtain the corresponding recognition result matrix P; step S3, through the formula F =A*P to get the voting result matrix F; step S4, calculate the maximum voting value in the voting result matrix; step S5, automatically mark the visibility level corresponding to the maximum voting value to the fog picture to be marked; step S6, repeat the above steps S2~ S5. Automatically mark the visibility level of all the fog images to be marked in the fog image set to be marked. The invention avoids the problem of inconsistent labeling results caused by manual labeling, and has higher labeling efficiency.

Description

technical field [0001] The invention belongs to the technical field of image recognition, in particular to a method for automatically marking the visibility level of a highway fog picture. Background technique [0002] The field of deep learning has developed rapidly recently, especially in the application of image recognition. The fog recognition based on highway monitoring images has also begun to apply deep learning technology. The key to deep learning technology lies in the collection and labeling of a large number of training sample pictures, and the difficulty of expressway cloud recognition lies in the labeling of the visibility level of sample pictures. The following reasons make it difficult to match the visibility value of the visibility detector with the monitoring image: [0003] 1. The deployment density of visibility detectors is low [0004] On average, one set of highway visibility detectors is about 10 kilometers away, and the number of intact equipment is ...

AI Technical Summary

Problems solved by technology

[0027] The purpose of the present invention is to overcome the deficiencies in the prior art, to provide an automatic labeling method for the visibility level of expressway group fog pictures, to avoid the problem of inconsistent labeling results caused by manual labeling, to use multiple group fog detection algorithms to vote, to select The visibility grade with the most votes is used as the marking result, realizing scientific automatic labeling

Images