A Model of Faint Target Temperature Estimation Based on Dual-band Infrared.

When detected far away, the temperature feature is important for target recognition. Based on infrared dual-band temperature measurement principle, target temperature feature extraction and confidence estimation model are established. The static experiments show that when the image SNR reaches 6, the feature estimation accuracy is 5K. The field test experiments show that when the image SNR reaches 6, the feature estimation accuracy is 10K.