The reasons and methods for generating and suppressing mid-spatial frequency (MSF) errors were deeply studied, and a new method of combining multiple small tools was used to smooth out MSF errors. First, an elastic tool was employed to pre-polish the aspheric surface after grinding, aiming at fast removing sub-surface damage as well as correcting surface figure. Then a second tool-self-designed smoothing tool -was mainly used to smooth the surface, which was followed by an air bonnet tool to continue figure correction until Zernike residual was not convergent. The effective combination of the three small tools was not only able to avoid the high slope fabrication difficulty of the aspheric surface, but also to suppress the MSF errors. The surface was tested by CGH, which can ensure the accuracy of the surface. This method was successfully used to polish a 150mm asphere with a maximum departure of 0.26mm. After five iterations, the surface accuracy converged to 4.05nm RMS. The result shows that this method can realize the valid polishing of high-accuracy aspheric surface.
Geiger-mode detectors have single photon sensitivity and picoseconds timing resolution, which make it a good candidate for low light level ranging applications, especially in the case of flash three dimensional imaging applications where the received laser power is extremely limited. Another advantage of Geiger-mode APD is their capability of large output current which can drive CMOS timing circuit directly, which means that larger format focal plane arrays can be easily fabricated using the mature CMOS technology. However Geiger-mode detector based FPAs can only measure the range information of a scene but not the reflectivity. Reflectivity is a major characteristic which can help target classification and identification. According to Poisson statistic nature, detection probability is tightly connected to the incident number of photon. Employing this relation, a signal intensity estimation method based on probability inversion is proposed. Instead of measuring intensity directly, several detections are conducted, then the detection probability is obtained and the intensity is estimated using this method. The relation between the estimator’s accuracy, measuring range and number of detections are discussed based on statistical theory. Finally Monte-Carlo simulation is conducted to verify the correctness of this theory. Using 100 times of detection, signal intensity equal to 4.6 photons per detection can be measured using this method. With slight modification of measuring strategy, intensity information can be obtained using current Geiger-mode detector based FPAs, which can enrich the information acquired and broaden the application field of current technology.