Lidar detects targets by actively emitting and receiving laser signals, and obtains information such as the target's distance, speed, and direction from the reflected echo. FMCW lidar is a ranging technology based on frequency modulated continuous wave signals, combining the advantages of frequency modulated continuous wave ranging and laser detection. This technology transmits a frequency modulated continuous wave to the target, mixes the signal light with the intrinsic light, and uses the beat frequency signal generated by laser interference to calculate the distance and speed of the target, thereby achieving high-precision distance and speed measurement. It has the characteristics of integrated distance and speed measurement and strong anti-interference ability. However, during laser transmission, depolarization effects may occur due to atmospheric turbulence or reflection from the target surface, resulting in polarization mismatch between the echo light and the intrinsic light. This polarization mismatch can significantly reduce the heterodyne efficiency, signal-to-noise ratio, maximum detection distance, and detection accuracy of the system. Therefore, in order to achieve high-precision long-distance detection, research on the polarization characteristics of FMCW lidar echoes has important academic significance and application value.In actual detection of FMCW laser radar, the polarization matching degree between signal light and intrinsic light directly determines the heterodyne efficiency of coherent detection, which in turn affects the signal-to-noise ratio, detection distance and accuracy of the FMCW laser radar system. Since the polarization state of the intrinsic light remains stable when it is transmitted in the polarization-maintaining fiber, and the polarization state of the signal light changes due to the scattering of the detection target surface when it is received by the optical system after undergoing a complex atmospheric propagation process. Therefore, it is particularly important to analyze the polarization characteristics of the echo in order to better understand its impact on the performance of the FMCW laser radar system. Based on the optimization and improvement of the traditional FMCW laser radar detection principle, a new detection system based on polarization orthogonal demodulation is proposed. A 1/4 wave plate is added to the incident end of the signal light, and the different polarization states of the signal light are simulated by changing the fast axis angle of the wave plate. The intensity of the balanced detector output signal is analyzed and verified.In view of the polarization mismatch between the echo light and the intrinsic light caused by the change of polarization state caused by the surface scattering of the detection target, this paper analyzes and studies the detection performance of FMCW laser radar based on polarization orthogonal demodulation from the perspective of target detection and recognition, effectively solving the problems of weakened detection capability and limited detection distance caused by polarization mismatch. The experimental results show that under the same test environment, compared with the typical detection method, the side mode suppression ratio of the six detection targets (reflector, quadcopter UAV, bird feather, green luo, marble, plastic bag) is improved by 11.85%, 11.98%, 13.46%, 15.39%, 17.12%, and 18.29% respectively. Further experiments show that by adjusting the wave plate angle (22.5°, 45°, 60°), when the wave plate angle is 45°, the ranging standard deviation reaches the minimum value of 0.012 m, and the side mode suppression ratio can be increased to 16.931 dB. Through the study of the echo polarization state, further theoretical support is provided for the application of FMCW laser radar in target recognition.This article combines polarization information with FMCW laser radar technology to provide theoretical basis for the polarization characteristics of echoes, expand their applications in military, autonomous driving, and industrial detection fields, improve target recognition, low reflectivity target detection, and environmental perception capabilities, solve the recognition limitations of traditional laser radar in complex environments, and provide theoretical support for high-precision target recognition and system optimization design.