In the highway traffic environment, radar video echo signals have the characteristics of large data volume, high noise interference, and fast target dynamic changes. If sparse representation and signal compression are not carried out in a timely manner, redundant information will mask the key features of the target, leading to a decrease in the joint detection performance of radar video echo signals. Research on a joint detection method for radar video echo signals based on quantized compressive sensing. By combining quantization technology with compressive sensing theory, sparse representation and quantization encoding are used to compress radar echo signals, which can effectively separate traffic targets (such as vehicles and pedestrians) from background noise and reduce a large amount of redundant information. And a joint detection algorithm is used to reconstruct and extract the compressed signal, in order to achieve joint detection of radar video echo signals. The experimental results show that the signal fluctuation amplitude of this method is consistent with the experimental indicators, ranging from -2dB to 1.5dB. Moreover, the signal fluctuation amplitude is also consistent with the experimental indicators when the signal frequency is between -10MHz and 10MHz, indicating that the detection results using this method are accurate. At a low signal-to-noise ratio of 18dB, the longest processing time is only 10s, indicating that this method has high real-time processing efficiency. The detection accuracy reaches 96%, which can effectively improve the accuracy and efficiency of object detection in highway traffic scenes.