What dFace Does
dFace helps organizations standardize, assess, and govern ID photos across both live capture and existing photo libraries.
It provides two core workflows:
Real-time processing and compliance check for individual photos
Batch processing and compliance assessment for stored and legacy photos
Together, these workflows help customers improve operational consistency at capture and increase the usability of historical photo data.
Workflow 1: Real-Time Processing and Compliance Check
For individual photos captured at service counters, self-service kiosks, mobile devices, or uploaded from smartphones, dFace provides a real-time workflow for standardized processing and compliance checking. Through smart cropping, foreground–background separation, and lighting and color correction, the system normalizes photos captured in complex environments while preserving the authenticity of the subject’s facial features. This significantly improves usability and consistency across devices.

Figure 3. Standard ID Photo Processing Workflow
In the standardized processing stage, dFace addresses common background-related issues in outdoor and open capture environments through subject extraction and background normalization. For photos with cluttered backgrounds, uncontrolled surroundings, or complex subject-background boundaries, the system can effectively separate the subject from the original scene and provide a consistent basis for standardized output.

Figure 4. Foreground-Background Separation
At the same time, for imaging issues such as harsh lighting, side lighting, local highlights, and color casts, the system applies controlled exposure balancing and color correction. For subjects whose images show localized specular highlights under strong lighting conditions, dFace also supports highlight reduction to minimize the impact of such reflections on image readability and downstream processing.

Figure 5. Highlight Removal
After processing, dFace automatically performs a comprehensive compliance check on the photo, covering key factors such as framing, background, lighting, pose, occlusion, and overall image quality. These checks are aligned with international standards, including ISO/IEC 39794-5, for portrait image quality and usability (see technical documentation for details), helping ensure that each portrait image meets relevant quality requirements.
Based on these checks, the system can provide clear outcomes and supporting reasons, such as:
whether the photo is compliant;
whether it can proceed to the next business step;
whether it can be brought into compliance through automatic correction; or
whether manual review or recapture is required.

Figure 6. ID Photo Compliance Check
This workflow can be flexibly deployed across desktop capture, mobile capture, and smartphone upload scenarios. It supports edge deployment, multi-platform compatibility, and multilingual prompts, enabling adaptation to cross-regional projects across different countries and device environments. dFace can also be integrated with EMPTECH capture devices to support an end-to-end capture solution, shortening integration cycles and reducing implementation complexity.

Figure 7. Deployment Options and Compatible Platforms
This workflow has already been validated in multiple national ID document capture projects. In one passport project, for example, dFace was deployed as a core part of the photo capture workflow, covering both service counters and distributed capture sites. With real-time compliance checking and automated assessment, the system helped the issuing authority maintain consistent quality standards in a high-throughput environment. It also reduced rework and manual review costs associated with ID photo quality issues while helping keep output photos consistent and compliant.

Figure 8. Application Illustration
Workflow 2: Batch Processing and Compliance Assessment
To address the challenges of legacy photo libraries and stored photos from multiple sources, dFace provides a workflow for batch standardization and compliance assessment. The system can apply consistent smart cropping, background normalization, and brightness and color balancing to thousands or even millions of existing photos. It can then perform batch quality and compliance assessment to improve the consistency and usability of historical photo data.
After batch processing, the system outputs not only standardized image files, but also structured issue reports and batch statistics indicating which images can be used directly for document production, which require manual review, and which must be recaptured. This workflow is well suited to centralized platform deployment and integration with higher-level business systems, helping customers manage and standardize existing photo data efficiently without launching large-scale recapture programs.

Figure 9. Batch Processing Workflow
In one national project, dFace’s batch processing capability played a key role in addressing large volumes of legacy photos from multiple sources with highly uneven quality. The system processed and standardized millions of stored photos, converting as many as possible into consistent and compliant ID photos despite differences in lighting, background, and capture devices. This provided more reliable input for downstream processes. Batch quality and compliance assessment then provided clear and actionable results for subsequent card issuance and data governance, significantly improving the reusability of historical data.
Through these two workflows, dFace extends ID photo quality control from a single function into a practical operational capability. On the one hand, it supports real-time capture scenarios such as service counters and mobile channels by providing immediate processing outcomes and compliance decisions for individual photos. On the other hand, it supports legacy photo governance by enabling batch correction, compliance assessment, and standardized output for stored photos. Together, these workflows support one goal: to provide customers with usable, compliant, and clearly assessable ID photos across different sources, environments, and project settings.
Last updated
