Research on the application
of ultra-small scale color IR Aerial Photography in investigation of
renewable natural resources and environment Yu
Xianping Department of Remote Sensing of Environment and Resources, Research Institute of Forest Resource Information, Chinese Academy of Forestry Abstract The paper introduces the present situation of application of ultra-small scale color IR aerial photography, and analyses its spectral resolution, spatial resolution, integrated resolution, thematic presentation capability and cost, to investigate the technical and economic properties of its application in survey of environment and renewable natural resources. The study shows that the application of ultra-small color IR aerial photography has a bright prospect. Aerial photography has got more and more wide use due to the rapid development of social economy and due to the increase of application demand. However, there is now a tendency worth discussion in the application of aerial photography, i.e. the scale becomes more and more large, which evidently needs more cost. An unnegligible principle of aerial photography is that the accuracy needed for research and application is the key factor for selecting the appropriate scale. In fact, in most cases, the abundant information of airphotos is far beyond enough exploitation and utilization in application. Based on the above-mentioned reason, the paper mainly discusses the application of ultra-small scale color IR aerial photography, and evaluates its applicability and economical benefit. This would help to make a proper choice of the scale of the airphotos, and to understand the potentials of such ultra-small scale photography, as well as to expend the application of such aerial remote sensing imagery. The ultra-small scale aerial area photography at the scale of 1:130000 was taken in somewhere, Hebei Province during the implementation process of the key scientific project in the Seventh National Five-Year Plan, the Comprehensive Investigation of the "Three Norths Protective Forest" by Remote Sensing, and thus started the prelude for the application of ultra-small scale color IR aerial photography. At present, the aerial photography at the scale of less than 1:100000 is called ultra-small scale Aerial photography. The domestic and foreign references, and the application conditions shows that neither the information potential of the ultra-small scale color IR aerial photography has been systematically quantitatively measured, nor the application technology investigated, nor the application experiment in operational scale has been made. In general, its spreading prospect is still in a state of obscurity. This study has made an analysis with the photos taken in somewhere, Hebei Province as examples, which are color IR aerial photos of the scale 1:130000 to investigate the potential and prospect of its application. Spectral Resolution The color IR aerial photos of the scale 1:130000 taken is somewhere, hebei Province in 1986 have excellent visula interpretation results. The images possess sufficient clearness and chromatic saturation even after linear magnificaion of 25 times (the enlarged images have a scale of about 1:5000). On the ultra-small scale color IR aerial photos enlarged to the scale 1:5000, all the ground types and objects including various tree species can be identified according to their color and other image characteristic. Chinese pine, black locust, poplar, oak, apricot tree, orchard tree, shrubs, barren land, farmland, road, water body, can be identified. In addition, the shape and size of the smaller objects (e,g. residential houses, small ponds) can also be clearly distinguished. The density of various objects on the image is also an impotent indication to their spectral characteristics, and is significant for identification of various objects. The study showed that the density difference between objects on ultra-small scale color IR airphotos is obviously larger than that of corresponding objects on standard false color composites of space imagery including SPOT, LANDSAT-TM, MSS. It is evident that large density difference favors visual interpretation of objects. Therefore, different tree species and other objects that possess relative difficulty for visual interpretation cn be relatively easily distinguished according to the differences of their spectral characteristics on ultra-small scale color IR air -photos. Spatial Resolution Generally, the following formula is accepted for spatial resolution: G = Rf / H Where G----ground resolution (pair of line per meter) H----flight height (m) f-----focal length of the camera lens (mm) R---overall resolution of lens and film system (pair of line per milli metr) Now, the commercially applied advanced aerial cameras can have a spatial resolution as high as up to 150 pairs of line mm (e.g. RMK, MSK 4etc.). Color IR film also develops in the direction of high resolution (its resolution is higher than 100 pairs of line/mm). As a consequence, the overall resolution of the lens/film system can be higher than 100 pairs of line/mm. In addition, color IR air photography possesses a good capability to penetrate fog, and the elimination of scattered blue light improves film contrast, and the difference of reflectance of vegetation of scattered blue light improves film contrast, and the difference of reflectance of vegetation in phographic IR section is usually larger, The higher contrast ratio leads to better spatial resolution. Therefore, color IR airphotos (especially film) have quite high spatial resolution. In somewhere, Hebei Province, the RC-10 aerial camera with f=88mm was used to take the color IR photographs of scale 1:1300000. Suppose the overall resolution of the system to be 100 pairs of line per mm, the deduced ground spatial resolution of the airphotos would be G = Rf/H = (100 x 88) / 11434.8 = 0.7692 (pairs of lines
per m) Theoretically, the minimum object width that could be distinguished on the photos is 0.65 meter. It is known through measurement on the enlarged ultra-small scale color IR air photos and through comparison with ground measurement, that the measurement on the photos has a relative error of only 3.2% and the spatial resolution exceeds one meter. These results coincide with theoretical calculation. The shape, size shadow and the relative of objects could be interpretated on the enlarged images of ultra-small scale color IR aerial photos through stereoscopic instrument (e.g. and analytical plotter,) thus three-dimensions spatial information could be obtained, which could be called the three-dimensions spatial resolution. Such property of the photographs is of important elements in forest inventory. This is extremely interesting in the field of forestry remote sensing, since they have seen a new way of forest resources inventory that is highly effective and of low cost. Examples of measuring some isolated trees indicated that the accuracy of three dimensional measuration of ultra-small scale color IR aerial photos might be rather high, and the relative error could be less than 5%. The extra-small scale color IR aerial photos that a possess a good three-dimensional resolution greatly strengthen its characteristics of interpretability and measuration, and spead a wide prospect for its application The Integrated Resolution The spectral resolution and spatial resolution of remote sensing images are inseparable, and to comprehensively analyse them would help image interpretation. According to the researches of J.a. Howard and D.C.Schwaar, maize ridge of about 0.7m ground width colud be identified by its integrated effect of the color IR aerial photos of the scale 1:120000, this could be called integrated resolution. Discussion on application technology of ultra-small scale color IR aerial photos
The ultra-small color IR aerial photos that possess the above mentioned high resolution has a capability to distinguish the following objects: Mature forests, forest stands, shrub forests, patch of broadleaf-grass, patch of the grass family, aquatic vegetations scattered forest swamp, vegetation on flooded plains, rock outcrops, sand dunes and winderosion lands, large stracts of naked earth, dry earth, moist earth, broooks, hydrologic types, small stream of grade two and three, landscape structure, landscape tones, landscape limits, landscape morphology, drainage condition, area of sources of water, severe soil erosion, forest fire slash, fallen wood , litter layer, cutover area, forest boundary, firing of forest and farmalnad, finge of forest and wasteland, farmland vagetation of landuse, desease and insect influence, desease damage, hudrologic conditions, natural open terrain, tree ages, treecanopy coverage, site conditions, forest stock volume, maturation degree of grasses, types and classes of gmsses, narrow paths, houses etc. Cost assement of ultra-small scale color IR aerial photography and prediction of its applicaton prospect Compared with normal middle scale color IR aerial photography, the cost of the ultra-small one could be reduce by 90%, and is only 0.25%. Its spatial resolution is 15 times higher than that of TM images, and its cost is only 2-4 times of the latter (in the same scale and same area) . Compared with SPOT image, the spatial resolution is 5 times higher, and the cost is only 56-100% of the latter. Therefore, the property-cost ratio of the ultra-small scale color IR aerial photography is very high. Its application is attractive both technically and cost effectively. Its application prospect is estimated to be very good. The application research of ultra-small scale color IR aerial photography conforms to the nee tendency of developemtnt of internatioal earial remote sensing. Its economic and technical sig. Nificance is also very great. The development it has obtained presages that the technology of aerial remote sensing and aerial survey would undergo important changes. The production application research of the ultra-small scale and high altitude color IR aerial photography has revealed the prelude of application of such photography in China to the aerial remote sensing investigation of renewable natural resources and environment, and showed a broad application prospet. Such developments have provided a theoretical basis fore the further development of comprethensive remote sensing investigation techniques of resources and environment, and would have broad application value. |