Abstract: Eddy current pulsed thermography (ECPT) and long pulse optical thermography (LPOT) as two emerging nondestructive testing and evaluation (NDT & E) techniques have been widely used for defect evaluation in many scenarios. This paper focuses on detecting subsurface defects in an aluminium plate by using both techniques. Principal component analysis and K-means clustering are used to enhance and extract artificial flat bottom hole (FBH) defects with different sizes. Results show that ECPT can detect 14 out of 18 FBHs, whereas LPOT can detect 12 out of 18 FBHs. Additionally, ECPT outperforms LPOT for detecting shallower FBHs and LPOT has an advantage in detecting FBHs with larger diameters.