The services of Payeshgaran Farasanjesh Tabiat Fars are organised into four lines: remote sensing, geographic information systems (GIS), UAV photogrammetry, and the design of smart spatial systems. Each line can be commissioned on its own, but the greatest value appears when the four are linked into a single chain that runs from raw data to an operational decision. This page describes the technical structure of each line, how they connect, a typical work cycle from order to handover, and the quality-control mechanism applied to deliverables. The aim is to let a technical buyer know, before ordering, the scope of work, the required inputs, the expected outputs, and the limitations of each line.
Four service lines and how they interlock
The first line is remote sensing, covering the acquisition and analysis of satellite and aerial imagery for wide-area coverage, periodic monitoring, and extraction of land-cover indicators. The second is GIS, which structures, stores, analyses, and publishes spatial data. The third is UAV photogrammetry, which produces high-detail data at local scale, and the fourth is the design of smart systems, which turns that data and analysis into a service usable by the end user.
The four lines meet at a shared data model. The output of remote sensing, namely cover and index layers, is stored in the same spatial framework as the output of photogrammetry, namely orthomosaic, elevation model, and vector layers. Layers from two sources with different accuracies therefore sit in one database and can be analysed together. This co-registration is the precondition for any multi-source analysis.
For the client, this connection means that a single analysis does not have to become a set of separate data purchases from several suppliers. The required inputs, the spatial reference, and the delivery formats are defined once, up front, and the accuracy gap between sources is declared explicitly and documented.
Remote sensing: from raw scene to information layer
In this line the sensor family is first matched to the client’s question: optical sensors for land cover and vegetation indices, thermal sensors for surface temperature, and synthetic aperture radar (SAR) for imaging independent of sunlight and with relative penetration under cloudy conditions. The choice is not governed by resolution alone; revisit interval, spectral and radiometric resolution, and the time window of the event also decide it.
Before analysis, the raw scene passes through a pre-processing chain: geometric correction and alignment to a reference, atmospheric correction to reduce scattering and absorption effects, cloud and shadow masking, and, where needed, merging of the panchromatic band with multispectral bands (pansharpening). The next step computes spectral indices such as NDVI, NDWI, NDBI, SAVI, and EVI, and runs supervised or unsupervised classification.
- Image acquisition and selection of sensor and time window suited to the phenomenon.
- Image analysis: index computation, classification, change detection, and time-series analysis.
- Output as raster and vector layers with metadata describing the method.
The achievable accuracy depends on sensor ground sample distance, the quality of atmospheric correction, and the quality of reference data used to train the classifier; these are recorded in the method report so the client can judge the accuracy in the context of their own application.
GIS: structuring, analysing, and publishing spatial data
The GIS line rests on four pillars: data model, spatial reference, database, and analysis. The data model is built on the three vector geometries (point, line, polygon) plus raster and attribute tables, and its design fixes the level of normalisation and the relationships between tables.
The spatial reference is kept identical across all stages; for data in Iran this is typically a UTM projection zone with its EPSG code on WGS84 or on a national datum, and transformations are executed with their parameters recorded. The database can be designed on PostgreSQL with PostGIS so that spatial and attribute queries run on one platform.
- Survey and preparation of spatial data and consolidation of the client’s existing data into GIS-ready form.
- Digitising and editing, definition of topology rules, and geometry validation.
- Spatial analysis: overlay, buffer, network analysis, and suitability modelling.
- Deployment of a spatial database with metadata and a unique identifier per feature.
For publication, data is served through OGC standard services such as WMS, WMTS, WFS, and WCS, and interchange formats include GeoJSON, GeoPackage, GeoTIFF, COG, Shapefile, KML, and DXF. Metadata is recorded against ISO 19115/19139 so that provenance and quality remain traceable.
UAV photogrammetry: accuracy at local scale
UAV photogrammetry suits areas that need accuracy higher than satellite imagery can provide. The method rests on aerial triangulation and bundle adjustment: from images with sufficient along-track and across-track overlap, matching points are extracted, camera positions and orientations are reconstructed, and a dense point cloud is produced by structure-from-motion and multi-view stereo (SfM/MVS).
Georeferencing is controlled with ground control points and RTK/PPK methods so the model aligns with real coordinates. Final product accuracy is normally stated as root-mean-square error (RMSE) expressed in ground sample distance (GSD) units; this figure is the basis for judging whether the data fits the application.
Deliverables from this line include the aerial orthomosaic, digital surface, terrain, and canopy height models (DSM/DTM/CHM), contour lines and breaklines, and volume estimation.
- Survey and preparation of accurate spatial data for the project area.
- Topographic mapping and volume estimation.
- Up-to-date mapping, parcel mapping, and aerial orthomosaics.
The limitations must be stated just as clearly: occlusion and shadowing, dense vegetation, water surfaces, and specular reflective surfaces all reduce the quality of the elevation model and the point cloud, and they are accounted for in flight planning and in the product report.
Designing smart spatial systems
This line turns data and analysis into a service for the end user. Its scope runs from WebGIS platforms and processing engines to systems that embed artificial intelligence. A typical architecture is built on a spatial database, a service layer, and a web or mobile interface.
On the intelligent side, semantic segmentation, object detection, and image classification models are applied, and their quality depends on the accuracy of the labelled training data; this point is made transparent with the client before a project starts. Model evaluation is reported with measures such as precision, recall, and intersection over union (IoU).
- Design of WebGIS platforms and remote-sensing systems.
- Design of intelligent processing engines for imagery and spatial data.
- Development of web interfaces, dashboards, and field applications.
Work cycle from order to handover
The cycle opens with a requirements meeting: the client’s question, the geographic extent, the required scale, the acceptable accuracy, and the delivery date are fixed. At the same stage the list of inputs expected from the client and those supplied by the company is made explicit.
Technical design and execution then follow: source data selection, flight planning or image ordering, running the processing and analysis, and building the final layers. At each step a version of the output is made available for review, so that a drift from the client’s expectation is corrected before the work ends.
The end of the cycle is devoted to delivering the product together with the method report and metadata and, where needed, training for use. In system projects this stage includes deployment, data migration, and initial support.
Quality control, metadata, and product handover
Quality control is applied at three levels: geometric accuracy, thematic accuracy, and completeness. Geometric accuracy is checked by comparing independent control points and computing the error, thematic accuracy by sample review against reference data, and completeness by comparing the delivery list with the contract extent.
Every product is delivered with metadata recording the spatial reference, date, data source, processing method, and known limitations. This record makes the results traceable and reproducible later.
- A written method report and delivery list.
- Data delivered in standard formats suited to the client’s system.
- Option of use training and knowledge transfer.
Summary
The four service lines, from image and field-data acquisition to spatial analysis and system building, form a continuous chain in which the output of each link is the input of the next. This continuity frees the client from coordinating several separate suppliers and secures a single spatial reference and data format.
A correct understanding of this structure helps the client define the scope realistically, match the accuracy expectation to the data source, and know the limitations of each method before ordering. Transparency in inputs, metadata, and the method report is an inseparable part of these services.
To begin, it is enough to establish the client’s question, the geographic extent, and the required accuracy level; on that basis the right combination of service lines, the deliverable outputs, and the work stages can be defined.