4.| RulesforJudging Aerial Haze and Smoke 
It is known that aerial haze and smoke showes large 
variety within the 
photographic season and aerial conditions which are 
highly responsible for degradation of fine details in 
picture. The flying staff decides in situ whether the 
atmospheric conditions may be tolerable from the 
viewpoint of Mission. The program takes into account 
the quotient of horizontal sight and flight height. The 
aerial conditions are good ifthe value ofthis quotient is 
higher. 
There is another way for judging haze conditions. It is 
known that the logE range of the film among others 
depends upon the haze conditions. It was observed 
that maritime air mass decreases visibility higher than 
continental ones and it decreases the the logE range 
of photographs and the contrast of fine details as well. 
Eight rules deal with the evaluation of haze conditions 
in this system. The emphasize is made on logE range 
calculated from the sensitometric and image densito- 
metric data, but the horizontal sight is interpreted too. 
In this program the logE range > = 1 is accepted as 
excellent value. In our test group which cosisted of 39 
films we have found only one film that reached this 
logE range. 
4.2 Rules for Evaluation of Camera Exposure. 
Although cameras with automatic exposure control 
are generally used today, the evaluation is necessary 
in order to refine the automation and cooperation 
between flying and laboratory staff. The exposure is 
evaluated from the sensitometric and image densito- 
metric data. Using the Dmin value of image density five 
rules deal with the interpretation of the camera expo- 
sure. 
4.3. Rules for Evaluation of Image Contrast, 
The photographic development is studied with the 
help of the comparison of test and control sensito- 
metry. Other important factor is the density range of 
the film which is optimal between 0.9-1.3 D. Further 
four rules interpret the density range of the photo- 
graphs. 
4.4, Rulesfor Evaluation of Micro Image. 
The ES is prepared for interpretation of Edge Gradient 
Analysis optionally. It is supposed that EGA was made 
with the help of a separate measuring program from 
which the data are introduced into this data base. The 
system ranks the films from the micro images separ- 
ately so that the ranking after micro and macro densi- 
tometry does not lead necessarily to the same result. 
The micro contrast and tone reproduction are two 
different needs as it is documented in APPENDIX 
Il. Eight further rules deal with the interpretationof de- 
cutance and Q factor. (Gerencser,M., Winkler,P. 
1988.) 
624 
4.5 RulesforRankinginto Quality Groups. 
The second group of rules ranks the aerial films into 
one of four quality classes following the ideas of natio- 
nalaerial photographic regulations. The general classi- 
fication is the last decision of the Report as it is shown 
in APPENDIX I. If the film got "unsuitable" qualification 
the decision is reasoned as it is documented in 
APPENDIX Il. 
5. EXPERIENCES WITH THE USE OF ES. 
The ES runs properly if the knowledge base is filled 
upompletely with true data. It assumes and needs dis- 
cipline in data collection. But this requirement is fami- 
liar for everybody who deal with data banks and PCs 
at all. Only the microphotometric data may be omitted. 
Although the lack of some descriptive data do not 
hinder the running of the system, the appearance of 
"No comment" does not increase the value of the Re- 
port. It is important that the agreement and coopera- 
tion oftechnical staff give further possibilities for usable 
description of aerial haze and for further evaluation of it 
too. In this case only restricted key words would be 
used. 
6. SUMMARY AND CONCLUSIONS. 
Aerial Photographic ES has been developed for IBM 
PC and Compatibles. The system is written under 
Borland's Turbo Prolog as a stand alone .EXE file 
which uses a further .DBA knowledge base file as well. 
The .DBA file is loaded onto the main file and used as 
internal data base. Taking into account the 640 KB 
memory of the AT286, not more than a thousands film 
can be stored at once. The development of external 
data base is also possible in Turbo Prolog with some- 
what longer access times. The system is capable for 
developing and deleting data base, which always pre- 
cedes the qualification. The qualification rules and 
reasonings are built into the software, the ranking 
followsthe national regulations. 
The application of ES in production needs the best 
cooperation with the flying and laboratory staff in order 
to get reliable data and restricted key word description 
of the phenomena in order to build them into the 
system. The freer interpretation of rules in the first part 
of the Report may stimulate the personnel for thinking 
inthe processes, refining them continuaiiy, as the Pro- 
log searches the solution relentness as well. 
7. REFERENCES 
Boberg,A.,1989,:A Survey of Aerial Image Quality As- 
sessment Methods. Photogammetriska 
meddelanden,No.54, The Royal Institute of Techno- 
logy, Stockholm. 
Gerencser,M., Winkler,P.,1988,Computer Assisted 
Measuring System for Edge Gradient Analysis. (In 
Hungarian), p.443-447, Geodezia es Kartografia, Bu- 
dapest. 
Smith,P.,1988,:Expert System Development in Prolog 
andTurboProlog, Sigma Press, Wilmslow U.K.