近景动态摄影测量中应用工业相机拍摄图像时,由于多种复杂因素会产生图像边缘灰度衰减的现象,影响目标点识别和像点中心提取精度。通过研究一种针对广角全画幅工业相机的图像灰度补偿方法,以余弦辐射体作为辐射模型,将图像质量劣化因素分为离轴辐射灰度衰减和未知因素灰度衰减并分别补偿。以扩展的误差函数对未知因素灰度衰减模型进行近似,并通过数值方法求解模型参数,得到同图像灰度分布无关的补偿系数矩阵,该方法成功补偿了工业相机的灰度衰减。试验表明,在3.2m×2.5m平面测量环境中,采用AVT GE4900工业相机装配Nikon35毫米定焦镜头进行摄影测量,经补偿后目标点的空间坐标的标准误差、平均误差和最大误差均有近1个数量级的降低,基本满足了该算法对于摄影测量系统精度的改善要求。
When measured by close-range photogrammetry method by using industrial camera, many complex factors will exert an influence that image fringe grayscale attenuation, this will affect the process of target point recognition and ex- traction of image point center. Grayscale compensation method according to the wide angle full frame industrial camera is re- searched and designed. The research based on cosine radiator, split the factors of the image quality degradation into off-axis radiation grayscale attenuation and grayscale attenuation by unknown factors, and compensate separately. Using extend Gauss error function to approximate the model of grayscale attenuation by unknown factors, and solve the parameters by nu- merical computation method, and return the compensation matrix which is independented with the image grayscale distribu- tion. The method can solve the degradation problem successfully. The experiments show that the measurement system' s ac- curacy which carried AVT GE4900 industrial camera fit with Nikon 35mm focus fixed scene is improved by the compensa- tion method,the standard error,the average error and the max error of the space coordinate of objective points has been re- duced by the method in one order of magnitude, and it proves that the method basically achieved the purpose of improving the accuracy of the close-range photogrammetry system.