The earth′s heat radiation, as one of the unknown parameters between the space target and ground-based telescopes infrared imaging terminal, will reduce the traditional dual-band pyrometry forward accuracy. To improve the accuracy of target temperature estimation, an estimation function based on the measured number of electrons is established, and temperature inversed solution model for solving the maximum likehood function is obtained; dual-band forward solution method based on the number of electrons is derived, and the simulation and analysis of the algorithm is performed. The experimental results indicate that the target temperature inversion precision has a close relationship with the imaging detector signal-to-noise ratio, the earth′s thermal radiation estimation accuracy and emissivity difference between bands; when the signal-to-noise ratio is relatively low, the dual-band pyrometry is in no solution; when the signal-to-noise ratio is relatively high, the inversion temperature accuracy of the dual-band pyrometry and maximum likelihood function is quite; function method of maximum likelihood using constraints and limited memory quasi-Newton optimization algorithm can effectively avoid the objective function into a local minimum; maximum likelihood function method has a strong ability to resist noise interference, can expand the scope of object temperature solution, and with high precision; under the guarantee of signal-to-noise ratio, increasing the number of image-band in order to improve the accuracy of temperature inversion.