自动判读中运动点目标脱靶量预测的数图结合法

A Combined Method of Predicating Quantity of Missing Distance for Moving Dot Target in the Automatic Interpretation

运动点目标的脱靶量自动提取是摄影测量图像全自动判读的关键,它以运动点目标脱靶量的正确预测为前提。提出运动点目标脱靶量预测的新方法-数图结合法,将预测的下一采样时刻的观测角度扣除主轴的角度后转换为目标在图像平面内的预测脱靶量,并以实验检验其预测偏差和判读效率。结果表明:该方法预测误差小,且无目标丢失,可明显提高判读速度。

The method of extracting quantities of missing distance for moving dot targets is the key to the automatic Interpretation of photographic images, and it is directly based on the exact prediction of quantity of missing distance. A combined method to predict quantity of missing distance is presented. After the axis angle being subtracted from the predicted angle, the predicted angle is transformed to quantity of missing distance. Both the er- ror and the efficiency were examined. Experiments show that the new method, without losing targets, gives less position errors, and improves the efficiency quite a lot.