DIGITALNA ARHIVA ŠUMARSKOG LISTA
prilagođeno pretraživanje po punom tekstu
| ŠUMARSKI LIST 10-12/1976 str. 50 <-- 50 --> PDF |
SUMMARY The semi-ocular method for tree height estimation in a stereomodel 1. Photointerpreters can — according to their skill and experience — estimate the heights of forest trees and/or stands on the AP-s. When using this pure ocular method in order to avoid bias, they have to be very familiar with local forest conditions. More objective is the semi-ocular method for tree height estimation in the stereomodel, which is based on the determination of the vertical scale of the stereomodel. 2. The factor of the height exaggeration of the stereomodel K is the relation between the horizontal scale 1 : m and the vertical scale 1 : mv of the stereomodel (1). When the stereomodel photographed by the base-height ratio B/Z is viewed by the personal base-height ratio of the interpreter b,/d using a stereoscope with the magnification lx, the relation (2) is valid. There B is the air base, Z the flight height, b() the eye base, and d the normal viewing distance. Out of (2) mr could be calculated (3). Through introducing the average b0 = 65 [mm] and d = 250 [mm], furthermore the photo base b and Z = m f, the formula (5) is obtained. According to (5) the nomograph (Picture 1) was constructed. For b and m an average of the dimensions on the left and right photo is used. Supposing an average end lap p IJ/c = 60 and the contemporary picture sizes s the formulae (6), (6a) and (6b) are obtained. 3. If the stereoscopic height of the object in the stereomodel hs is estimated, usually in [mm], the height h in the nature in [m], is calculated by (7) or using a stereoscope with the magnification e by (8). The height in the nature could also be found by a »Stereoscopic heights table« (Table 1), which was constructed by using (6a) and (6b). 4. A trial in the use of the described method was performed by the authors. Three interpreters (A, B, C) estimated on three stereopairs in total 144 heights of trees and stands, whose heights were known from the ground measurements. The AP-s and ground measurements description, the results obtained and F-test performed are found in Tables 2 and 3. No significant difference among the interpreters was found, so the observations for each stereopair were cumulated. The difference between the variances for the stereopair combinations 1—2 and 1—3 was found significant when expressed in [mm] on the photo. This could be explained by the magnitude of hs, which is dependent on f and e. The greater hs are probably estimated less accurately. For all the stereomodels and interpreters the mean square errors were less than +2[m]. The average error for all the observations was about ± 1.6 [m]. 5. In deriving all the formulae, the normal vertical case and horizontal terrain were supposed. All the deviations from these conditions cause the smaller or greater systematic errors. The method described is suitable for the horizontal terrain, which, however is not so frequent in nature. If a maximal height error ± 0.03 h is allowed, a terrain with the maximal height difference of about 0.03 Z could be considered horizontal. If the height differences exceed this amount, the stereomodel is to be divided into several height zones and for each a new mv to be calculated, taking into consideration the relation between the dimensions involved, when they are at two height levels H and H; (11a, lib). Photointerpreters are to be trained on models with known heights. For persons whose b„ and d differ entirely from supposed ones, mv is to be calculated by (3). |