This article describes an approach of human face recognition based on wavelet transform from thermal IR images. The system consists of three steps: human thermal IR face image is preprocessed and the face region is cropped from the entire image; Haar wavelet is used to extract low frequency band; and image classification. They reported a recognition rate of 95%.

1.- Please complete the methodology stated in the abstract by mentioning the series matching technique used in the classification step.

Comments on introduction/background

2.- Update your reference 3 with more recent literature about infrared face recognition.I recommend the review "Infrared face recognition: A comprehensive review of methodologies and databases" by Shoja Ghiass et al.

3.- Please expand "IR" and "CCD" acronyms at the beggining of the introduction.

Comments on methodology

4.- Please include the demographic information of the 76 individuals involved.

Comments on data and results.

5.- Include the recognition rate (average obtained from the different datasets) of your method in the table 3 where you presented results from the other 10 methods so the comparison can be performed more clearly.

6.- Specify the units (pixels) used in table 1,

Comments on discussion and conclusions

7.- Mention your conclusion about if your method provides an advantage on the face recognition problem according to your results reported in table 3.

8.- Finally, include limitations about the sample size, demography, and acquisition process.