This paper presents direct time-of-flight image LiDAR to realize a reliable autopilot system, which consists of hardware architecture and processing algorithm based on field programmable logic array (FPGA). The processing algorithm consists of multi-channel time-to-digital converters (TDCs), equivalent sampling technology consisting of equivalent sampling circuit (ESC) and amplitude fitting algorithm (AFA), correlation slope fitting algorithm (CSFA), and stray light suppression algorithm (SLSA). The TDCs are used to measure the flight time. The ESC is used to record waveforms based on time discrimination circuits, dynamic threshold adjustment circuits, and TDCs. The AFA is used to achieve the amplitude of the echo signal whether it is saturated or not. The CSFA based on the Gauss function is used to eliminate walk error, which does not need to calibrate the compensation coefficient in advance. And the stary light suppression algorithm based on laser pulse signal modulation technology is used to suppress the interference of background light and interference within different LiDAR systems. The LiDAR system can achieve 100 m measurement range with the laser peak power of 120 W. A measurement precision of ~+/−15 mm was achieved with non-cooperative targets when the measurement range was less than 10m. And the measured relative precision was less than 1% for a distance up to 100 m with a 100-klux direct sunlight.