Method For Imaged Target Ranging Based On Line Segment Features

It invested into the passive ranging method based on the imaged features from a monocular imaging system. A ranging method was presented for imaged target ranging based on line segment features. The principle of the algorithm is to realize a three point match between two images in each adjacent frame, which is Scale Invariant Feature Transform (SIFT) direction guided Perspective-three-Point (P3P) algorithm. Based on these three points, a virtual circle is settled. Then, a directed string from the circle is selected for the line segment feature. With the line segment features and other parameters, a distinctive ranging equation could be settled for the distance estimation. Relative researches indicate that ranging error is mainly determined by the precision of the characteristic linearity, so how to select a better line segment feature was discussed. Experimental result implies that this method could obtain a good result.