在自动目标识别系统中,目标探测起着重要作用.文中利用美国陆军实验室提出的特征级融合方法,应用于相干激光雷达强度像中目标中心点位置的确定.即通过局部最大灰度特征、对比度特征、平均梯度值特征和局部变化特征,最终确定激光雷达目标中心位置.经实验仿真结果表明,相对形心跟踪器确定中心点,该方法不易受到环境的影响而且发现目标中心位置准确性更高,提高相干激光雷达发现目标的准确性. In automatic target recognition system, target detection plays an important role.This paper uses the feature-level fusion method proposed by the US Army Laboratory to determine the position of target center in the intensity image of coherent lidar.According to the maximum local gray feature, Contrast feature, average gradient value and local variation feature, and ultimately determine the target center position of the laser radar.The simulation results show that the relative center of the tracer to determine the center point, the method is not subject to environmental impact and found that the accuracy of the target center position is more High, improve the accuracy of coherent laser radar to find the target.