Recent past witnessed the introduction of a few nonstationary thermal wave imaging modalities to overcome the shortfalls of conventional thermographic methods. Quadratic frequency modulated thermal wave imaging (QFMTWI) is one of these techniques which imposes a continuous frequency sweep containing more energy with low frequency thermal waves in a single experimentation cycle to facilitate entire depth probing of the test object, unlike its linear frequency counterpart. In order to extract fine subsurface features in this modality, the temporal thermal history of the stimulated object is captured using an infrared imager and post processing methods like phase analysis or pulse compression etc., have been carried out. But all these existing post processing methods are based on Fourier transform which provides limited frequency resolution based on the number of captured thermograms and their capturing rate. This inadequate frequency resolution subsequently limits the depth resolution capability of this modality and required to be enhanced.a proposed a method Orthonormal projection which give better performance than other post processing methods. Performance of Orthonormal projection based approach is demonstrated by considering defect detection, signal to noise ratio (SNR).