The apparent water depth obtained by photogrammetric measurement appears to be shallower than true depth. This is because the reflected sun-rays from the sea-bottom refract at the water's surface.
In order to convert the water depth observed by aerial photographs to true depth, it is necessary to obtain correction factors. Equation (2), derived by G.C. Twinkel, determines the correction factors; equation (3), derived by J.G. Meijer, transforms the correction factors of equation (2) so that they may be used for numerical calculation by digital computer. In this study, the computer programs based on the above equations were successful. (During the study, A-7 plotting equipment was used for analyzing the aerial photographs.)
The two areas covered by the aerial photographs were: 1. off-shore, near Omaezaki Port, Shizuoka Prefecture; 2. Amatsu Port, Chiba Prefecture. Both areas are located in Japan.
The water depths observed from the aerial photographs were corrected to true depths by digital computer. The calculated true depths were compared with the water depth results recorded by echo-sounding. The results of this comparison were as follow:
1) The apparent water depths obtained by photographs measured deeper than those obtained by echo-soundings.
2) The mean differences of the water depths obtained by aerial photographs and echosoundings were 13.9cm in the Omaezaki area (within 500m from the shoreline) and 5.4cm at Amatsu Port.
3) The mean water depths by echo-soundings were 3.24m off-shore Omaezaki and 0.62m at Amatsu Port. Consequently, the relative errors were within a ±30% maximum range (and were mostly a ±10% range) .
4) As relative orientation of the stereo model of Amatsu Port was complete, we were able to carry out statistical test of the differences between the independent samples. Results showed that the distinct differences between these two methods of measurement were not recognized.
Conclusion; The application of photogrammetry to water depth measurement is practicable and may prove valuable in near-shore and/or reef zone areas where it is difficult to reach such arenas by observation boat.

View full abstract

Some environmental impacts were analyzed by the data handling way on quantitative manual working system of multispectral air photos.
1) Spectral reflectance properties of trees.—
The trees having the following characters were delicate to the waste gases; a) spectral reflectance is reduced, b) but, increased relatively in the near infrared region, c) also, the hi-band ratios (IR₁/G, IR₁/R) are high.
2) Contamination index of environment by air pollution.—
The following trees are primarily attacked by the waste gases; Zelkova, Cherry, Celtis, Sophora. The degree of the contamination is effectively measured by the hi-band ratio of their spectral reflectance.
3) Tree vitality and spontaneous combustion of trashes.—
It is possible to differentiate the vitality of the trees, which were not recognized any changes by appearance, with the bi-band ratios of spectral reflectance.
4) Natural slope and tree vitality, in the mountainous region.—
The natural slope being over 20 degrees, the bi-band ratios of the trees growing on the slope decrease apparently, and this slope agrees with one which the slope fall increases distinctly.
5) Tourist resort capacity viewing out of the vitality of green plants.—
It was estimated ecologically by the environmental condition affecting to the tree vitality.
6) Spectral features of sliding land.—
It is inferred from the spectral reflectance of the sliding land that the water content in the upper stratum, in which the landslide broke out, was high as compared with the middle.

View full abstract

This paper describes the method to determine discriminative mesh zones of four categolies by tone arguments on composed color photographies which are made by malti-spectral images, and the difference of accuracy between the test field and the training field.
The accuracy of each categoly is computed from frequency of discriminative mesh zone. Some essential particulars of the result of this research are as follows.
1. The apparent accuracy on the automatic processing of the discriminative mesh zone is about 83.5% and true accuracy is about 70.3%.
2. The standard of the discrimination of the test field data is difficult to use in the training field.
3. Using the discriminative mesh zone, uneven data of each categoly does not have appraisement.

View full abstract