Face Recognition Method Under Complex Light Conditions in Coal Mines

Fig. 2. Comparison before and after low-frequency signal equalization processing. (a) Original image and corresponding histogram; (b) image effect and corresponding histogram after low-frequency equalization processing

Download full size

Fig. 3. High-frequency processingeffects of downhole face images by different algorithms. (a) Original image; (b) traditional wavelet algorithm; (c) Log_Gabor algorithm; (d) bilateral filtered multiscale Retinex algorithm;(e) algorithm in Ref.[16]; (f) proposed algorithm

Download full size

Fig. 4. Partial sampling images in experiment

Download full size

Fig. 5. Post-texture feature maps described by LBP

Download full size

Table 1. Parameter setting in experiment
View table
Table 1. Parameter setting in experiment
Name Parameter Value
Wavelet fuzzyprocessing ScalesDirection 46
LBP description Neighborhoodsizes ALBP12, 1.5
LBP imagesegmentation Size 13×13

Table 2. Recognition rates by different recognition methods%

View table

Table 2. Recognition rates by different recognition methods%

Method	Recognition rate
Method	Subset 1	Subset 2	Subset 3	Average value
Method in Ref. [3]	75.00	66.67	66.67	69.45
Method in Ref. [4]	83.33	75.00	91.67	83.33
Method in Ref. [5]	75.00	66.67	75.00	72.22
Method in Ref. [13]	83.33	83.33	91.67	86.11
Proposed method	91.67	91.67	100.00	94.45

Tools

Get Citation

Copy Citation Text

Yuehua Huo, Weiqiang Fan. Face Recognition Method Under Complex Light Conditions in Coal Mines[J]. Laser & Optoelectronics Progress, 2019, 56(1): 011003

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites