Deceti.jpg (13851 bytes)

Structural elements of improving Photointerpretation and Remote Sensing Methodology

10.1 Introduction
10.2 Factors of Photointerpretation - Remote Sensing possibilities

10.1 Introduction

The structural elements of the photointerpretation and remote sensing methodology are:

a. Detection ability
b. Spatial resolution
c. Spectral resolution
d. Recognition ability
e. Correlation ability and
f. Evaluation ability.

In more detail:

a. Detection ability, is the ability of a remote sensing system, organic or not, to sense, detect and record the appearance or absence of an element/object/characteristic of the natural and built environment, even when its identity cannot be immediately determined.

b. Spatial resolution is the ability of a remote sensing system, organic or not, to distinguish two objects, in the minimum possible distance from each other, as discrete in their specific remotely sensed image.

Therefore, a remotely sensed image is characterized by its resolution, which is related to the resolving power of the remote sensing system as well as to a series of other factors concerning the specific acquisition conditions and to the characteristics of the scene’s objects.
Resolution cannot be considered as an objective measure, as it depends on the characteristics of:

BULLET.JPG (677 bytes) vision,
BULLET.JPG (677 bytes) critical analysis and
BULLET.JPG (677 bytes) perception ability of the photointerpreter,

in relation to the dimensions, shapes, structures, contrast, etc of the adjacent objects.

To overcome problems of resolution subjectivity we may use the Modulation Transfer Function, which expresses the overall formation/transfer functioning of the specific remote sensing system.
In order to calculate the dimensions of the minimum identifiable element/object/characteristic on an image, the scale must be first known.
We may define the angular value ã (in mrad) of the resolution of a remote sensing system which can discern 4 line pairs/centimeter, from a distance of 12.5m, as:

ã = 1.25mm / 12,500mm = 0.1mrad

c. Spectral resolution, expressing the interaction of electromagnetic radiation with the element/object/characteristic, results in the way it is presented on the image by the remote sensing system (spectral signature).

d. Recognition ability, is the ability to identify an element/object/characteristic by the way it appears on an image. It should be noted that it is possible that an element/object/characteristic of the natural and built environment cannot be recognised despite the fact that it can be detected and discerned form its neighboring ones in a remotely sensed image.

e. Correlation ability, is defined as the ability of a remote sensing system, organic or not, or of a combination of remote sensing systems, to:

BULLET.JPG (677 bytes) perceive and properly record point, linear, surface and spatial elements, appearances, arrangements, patterns and characteristics in their geometrical and/or spectral and space-temporal dimension and

BULLET.JPG (677 bytes) secure the conditions for their systematic correlation with relevant recordings on one or more remotely sensed images in an organic or automated way, using internal or other appropriate procedures.

f. Evaluation ability, in the photointerpretation/remote sensing methodology, is the ability of a remote sensing system, organic or not, or of a combination of remote sensing systems to estimate, evaluate and weigh the significance of information derived from remotely sensed images.

Based on the above, improvements on the photointerpretation and remote sensing methodology may be achieved by optimizing each one or combinations of these five structural elements.

 
10.2 Factors of Photointerpretation - Remote Sensing possibilities improvement

TOP

A basic factor for the best exploitation of remote sensing possibilities is, of course, the systematic and orderly analysis of each remotely sensed image’s “useful” information.
In order to achieve the best possible results by exploiting the possibilities of the remote sensing methods/techniques used, the photointerpreter has to be:
BULLET.JPG (677 bytes) able to handle successfully the powers of reasoning,
BULLET.JPG (677 bytes) an expert on the scientific field of the specific application (in order to properly plan and to select, in each case, the appropriate sensors, the most reliable and effective processing techniques and in order to organise the appropriate ground surveys,
BULLET.JPG (677 bytes) of wide general culture, (in order to use it, with any other bibliographical, statistical, meteorological, cartographic etc. data, as an infrastructure of the analysis process),
BULLET.JPG (677 bytes) of satisfactory vision and stereoscopic conception, (so as to work out all the identifying image characteristics of an image/a pair of images: tone/colour, shape, form, texture, pattern, shadow, sensation 3D, topographic location, relation to the local environment),
BULLET.JPG (677 bytes) of important technical experience in the specific field of application, (so that the observation and data analysis process can be documented on it, with the help of photointerpretation’s comparative processing, of critical analysis and photointerpretation keys),
BULLET.JPG (677 bytes) familiarised with multi-disciplinary approach principles, methods and techniques and with the integrated approach of the subject under investigation.

a. Photointerpretation has to be done orderly, so as:
BULLET.JPG (677 bytes) to elaborate one question at a time (in order to avoid confusion, disorientation and consequently non-effectiveness), starting from the simplest one and continuing to the more complex ones (depending on the subject and the photointerpreter's specialisation and experience),
BULLET.JPG (677 bytes) to proceed from the analysis and the study of the general to the analysis and the study of the specific, first by elaborating, for example, the general models of the study area on LANDSAT, SPOT, MOS etc. remotely sensed images, aerial photo-mosaics and non conventional photographs of small scale, and then carrying out a detailed and extensive study of more specific models/characteristics on aerial photographs/images of big scale or on aerial photographs/images enlargements,
BULLET.JPG (677 bytes) to draw specific conclusions from the general models' study (Deductive Evaluation). For example, from the study of a region’s surface drainage network, we can deduct reliable information on the soil’s consistency and from the region’s geomorphologic structure, some information on the soil’s fertility,
BULLET.JPG (677 bytes) to complete conclusion deduction by implication, from the specific study to the extraction of a region’s general models. For example, from the shape of a cross-section and from the length of a gully, it could be possible to draw reliable conclusions concerning the wider area, where  such gullies systematically appear (Inductive Evaluation),
BULLET.JPG (677 bytes) to achieve successive exploitation of the photointerpretation deductions. For example, according to the data/models/characteristics already extracted and known, it is possible to acquire (with comparisons and parallel elaboration) new data,
BULLET.JPG (677 bytes) to always take into consideration, as a basic factor influencing photointerpretation’s possibilities, the quality and the acquisition conditions of aerial photographs/images, because sometimes data could be degraded or even disappear (or exaggerated on some occasions), as a result of time, season, weather conditions at the time of acquisition (for example, growing stages of a yearly plant),
BULLET.JPG (677 bytes) to continuously exploit the mental process of successive approaches/elimination, in order to complete the steps described above, so as to identify the relevant objects/appearances/characteristics, or to limit investigation to the exploration of a small group of possible alternative solutions,
BULLET.JPG (677 bytes) to take into consideration the interdependence of the natural and the socio-economic reality, their relationship, interdependence and interactions, and their change trends through time,

b. to systematically exploit the basic image characteristics (shape, form, sensation of 3rd dimension, tone/shades of colour, texture, shadows, patterns, topographic location, relation to the local environment) from the direct and more familiar to the photointerpreter, to the indirect and more complex ones,

c. to suitably utilise the existing bibliographical, geographic, cartographic etc. material concerning the study area/object, in relation to the nature and volume of the qualitative information, available in each case.
This information relates to the process of its creation, formation and evaluation through time and the influence of the natural and built environment,

d. to fully exploit the most suitable technique of remote sensing processing, in relation to the available means, such as:

1. the possibility of stereoscopic vision of conventional infrared, coloured or the coloured-infrared aerial photographs,
2. the possibility of simultaneous viewing by two or more photointerpreters and comparative photointerpretation of aerial photographs/images or stereoscopic pairs in their analogue format, photographs and images of different scales, dates, seasons, acquisition time, enlargements, combinations of films/filters etc,
3. the possibility of parallel exploitation for the same area of images provided by the most convenient remote sensing/systems,
4. the possibility of analysis and automatic processing of the tone, the patterns and the texture of one or of a series of multi-spectral remotely sensed images,
5. the possibility of data exploitation to extract spectral signatures or other natural, chemical, biological properties of different objects,
6. the possibility of using the most appropriate interpretation keys in its analogue or digital format,

e. to always exploit the existing “libraries” of photointerpretation keys.

Reference:.Rokos, D. “Photointerpretation and Remote Sensing”, NTUA, 1979

 

TOP
ntualogo.GIF (15016 bytes)

National Technical University of Athens
Dept. of Rural & Surveying Engineering
Laboratory of Remote Sensing