An Aerial Photography - Aided Seaweed Bed Survey Techniques
Digital image processing
In this technique, digital image data covering the three band (blue, green and red) of positive color film is created and then the image is processed. Figure 2 shows the processing process.

Figure 2 Digital image processing
Maximum likelihood estimation is used for seaweed bed classification.
This technique is capable of giving a fairly objective classification of seaweed beds, yet is disadvantageous in terms of processing costs.
Analog image processing (masking) method
In this technique, images of seaweed distribution zones are directly printed onto films by a photographic processing technique that takes advantage of the blue, green, and red photographic intensity characteristics of positive color films. The basic seaweed bed classification principles are the same as those used in digital image processing.
Figure 3 shows the processing process.

Figure 3 Analog image processing
This technique enables us to obtain seaweed bed distribution charts equivalent to or better than those obtained by the digital image processing technique in terms of precision and quality at a relatively low cost. For this reason, this technique is expected to find much wide application in the future.
Summary
Table 2 lists a comparison of the different photo-analytic techniques.
Table 2 Comparison of Photo analytic techniques
The digital and analog image processing techniques are more objective than reading photos by eye, yet they may have some difficulty in discriminating between types of seaweed bed, depending on the pertinent photographic conditions a sea area conditions, in so far as the currently available classification methods are done only according to tone. To resolve this difficulty, it is necessary to develop techniques for the quantification of the texture and pattern photo reading keys used in reading photos by eye.
In view of these conditions, our view is that under the present circumstances it is most appropriate to use the method of reading photos by eye together with either analog or digital image processing.