The customer sends the photogrammetric images (e.g. acquired via UAS). The images are aligned and the MetricAware optimization is applied to the photogrammetric block. Once the process is completed, the optimized block is provided to the customer.

At present, optimized blocks are compatible with Metashape or COLMAP. The customer specifies the target software when submitting the images. Support for additional photogrammetric software is planned.

Importing the optimized block into Metashape or COLMAP is effortless. As needed, assistance for this step is included in the MetricAware optimization service.

MetricAware addresses the internal consistency of the photogrammetric block, independent of error metrics at control points or check points.

No. Functions such as Gradual Selection rely on manual tie-point filtering based on global criteria and user-defined thresholds. Although this approach can reduce reprojection error, achieving sub-pixel performance often requires extensive trial-and-error and the behavior is not necessarily monotonic. Beyond a certain level of pruning, additional filtering may even worsen the reprojection error. Furthermore, this process tends to erode tie-point redundancy and weaken network robustness, increasing the risk of sparse cloud degradation.

The non-equivalence with MetricAware lies in the quality of the results, as well as in the controllability, repeatability, and robustness of the process itself.

Open-source automatic pruning tools, often sophisticated in geometric and numerical terms, can achieve a marked reduction of the reprojection error by accepting a structural compromise in the photogrammetric block.

MetricAware explicitly controls structural impact during the optimization process, rather than treating it as a consequence of filtering.

The difference does not concern the ambition of the reprojection error reduction, but the structural cost accepted in order to achieve it.

No. MetricAware optimization provides measurable benefits even when the input imagery presents intrinsic limitations, by increasing the internal consistency of the photogrammetric block with respect to the available observations. However, the quality of the block remains constrained by the radiometric characteristics, acquisition geometry, and completeness of the image dataset.

Yes. Pre-calibrated cameras and RTK/PPK systems can reduce the initial uncertainty of the photogrammetric block. However, they do not eliminate the possibility that the block may still exhibit internal metric inconsistencies.

MetricAware improves the internal consistency of the block after image alignment, independently of the calibration or georeferencing strategies adopted.

MetricAware is designed as a controlled algorithmic optimization process, rather than as a generic executable software.

Although the optimization core is automatic and adaptive, its effective application depends on contextual decisions, parameterization, and stopping conditions that require appropriate oversight and quality assurance.

For this reason, MetricAware is offered as a service. This allows the optimization to be applied consistently and responsibly.

Pricing is determined on a quotation basis, as projects with similar data volumes may have different requirements.

The value of MetricAware optimization lies in delivering a clear and defensible metric result that downstream processing can reliably depend on.

To summarize: pricing reflects responsibility, not volume.

In contexts of specific technical, scientific, or institutional interest, forms of project collaboration may be considered as an alternative to a purely commercial relationship, subject to clear objectives and mutual value.

We don’t expect blind trust. Photogrammetric images are used exclusively for the execution of the requested activity and are not disclosed, transferred to third parties, or retained beyond what has been explicitly agreed.

Ileron has long-standing experience in the processing of industrial and technical data, including contexts with high informational sensitivity. Verifiable references are available upon request and include projects involving critical infrastructures, public administrations, large industrial groups, and defense-related technologies.

Concerns about sharing images for processing services are legitimate. Our responsibility is to ensure that data are handled strictly within the agreed scope and confidentiality constraints.