Recently, generation of the orthophoto using digital aerial photograph is performed for the maintenance of spatial information data for wide areas by middle scale. For orthophoto generation, digital elevation model (DEM) by the stereo matching methods is widely used. For the efficiency of orthophoto generation, mismatching should be removed and accuracy should be improved effectively in stereo matching process. With this motive, new method which is called “Reverse-Projection Method” to remove mismatching points on DEM generation is proposed. The most remarkable points of the method are its ability to detect mismatching points and improve accuracy of DEM using only neighbor stereo models on the strip or the next strip, and the effectiveness of the method is discussed in this paper.
Stable supply of high quality crop is an urgent requirement for rice cropping facilities in Japan. To cope with this demand, it is absolutely essential to evaluate rice leaf's chlorophyll content by measuring the SPAD values. But it requires lots of elaboration to measure the SPAD values for wide area in paddy. Therefore, the objective of this study is to resolve this problem by estimating the rice leaves SPAD values through remote sensing. Specifically, we made use of a trial observation equipment using hyperspectral sensor mounted on a radio-controlled helicopter, and calculated new indices LVIpure and LVImixture using the obtained data to estimate SPAD values. As a result, the relations between the SPAD value and the estimation index exhibits high correlation : R2=0.885 for LVIpure and R2=0.927 for LVImixture, which facilitated to obtain “the estimated SPAD value image” from the obtained data.