Three-dimensional imaging lidar can offer the range information and CCD camera can offer the color information of the target. These two information can strongly supplement each other, the fusion of them is becoming a research hot point. The major problem of fusing lidar data and camera data is the coordinate calibration between them. In consideration of the traits of lidar and camera, a new planar object calibration method was proposed, which solves extrinsic matrix using the normal of planar objects. The rotation matrix and transfer vector are solved separately, which can simplify the calibration process and highly improve the calibration precision. The improved extrinsic calibration method was applied on texture mapping and reverse mapping, which show that our algorithm can gives good results.