To improve the accuracy of infrared detection and realize synchronous detection of defect depth and size, a conjugate gradient recognition algorithm is integrated with pulsed thermography and pulsed phase thermography. Quantitative identification of infrared technology is achieved based on the phase and surface temperatures. The influence of these factors on the identification result is analyzed using numerical examples. The results show that without temperature errors, the defect is accurately identified based on the phase and surface temperature. The random temperature error decreases the accuracy of the identification result based on the phase and surface temperature. The uniform temperature error decreases the accuracy of the identification result based on the surface temperature. However, the uniform temperature measurement error does not affect the identification result based on the phase.