Automating image processing workflows

The challenge

A recurring weekly workflow required processing 30-50 images with a consistent cinematic visual style for eBook, audio, and video covers.

The manual workflow involved opening each image individually in Photoshop, applying multiple adjustments and visual effects, and exporting each image separately. This repetitive process typically required several hours of manual work every week.

The desired visual style focused on:

  • deep, rich colors
  • strong contrast
  • cinematic atmosphere
  • consistent visual identity
  • subtle variation between images

The manual workflow also increased the risk of inconsistent output caused by repetitive editing and production fatigue.

The solution

To optimise the workflow, I developed a Python-based automated image processing system using Visual Studio Code and AI-assisted development workflows with Claude AI.

The workflow combined:

  • automated batch processing
  • cinematic base image grading
  • randomised artistic overlays
  • structured export automation
  • consistent visual processing rules

The system processes entire folders automatically while maintaining a unified cinematic aesthetic across all images.

Workflow overview
Original images
        ↓
Base cinematic processing
        ↓
Randomized artistic filter
        ↓
High-quality export

The workflow uses a two-layer processing structure:

  • a consistent cinematic base layer applied to all images
  • randomised visual overlays and filters to create natural variation

This combination maintained consistency while avoiding repetitive-looking output across larger image sets.

Results and impact
Time savings
  • Manual workflow: 3-4 hours weekly
  • Automated workflow: 5 minutes weekly
  • Estimated efficiency improvement: 97-98%
Workflow improvements
  • consistent image processing across batches
  • reduced repetitive manual work
  • scalable processing regardless of image volume
  • predictable and repeatable output quality
Scalability

The system can process hundreds of images with minimal additional effort, making it suitable for larger publishing and content production workflows.

Manual workflowAutomated workflow
Images per batch30-5030-50
Time required3-4 hours5 minutes
ConsistencyVariableConsistent
ScalabilityLimitedHigh
Example

Before

After

Project Details

Workflow type
Automated image processing

Scale
30-50 Images weekly

Time reduction
3-4 hours → 5 minutes

Key focus
Automation
Batch processing
Visual consistency

Technologies
Python
PIL (Pillow)
NumPy
Visual Studio Code

Skills demonstrated
Workflow automation
Python scripting
Production optimisation
Batch processing