A novel state estimation method is proposed for target tracking in the boost stage using space-based infrared cameras (SBIRC) whose measurements are essentially corrupted by both Gaussian noise and quantization noise. As the quantization noise has non-Gaussian properties, conventional extended Kalman filtering (EKF) suffers from poor performance. The quantization noise of SBIRC is modelled using the mid-riser quantizer, which is usually adopted in the digital signal processing field. A novel minimum square error (MMSE) state estimation algorithm with quantized measurements, named the quantized extended Kalman filtering (QEKF), is then proposed. The time update is given based on first-order linearization of the nonlinearities, and the measurement update is derived based on the conditional mean estimate given the quantized measurements. As the multidimensional integrals in the measurement update derived doesn't have analytical solutions, a numerical integration method is proposed by combining Genz's transformation and quasi-Monte Carlo (QMC) method, which can avoid the curse of dimensionality. To further improve the tracking accuracy, quantized high-degree cubature Kalman filtering (QHCKF) is developed by integrating the fifth-degree cubature rule into the framework of the QEKF. Numerical simulation results illustrate the superiority of the proposed QEKF and QHCKF methods.