Skip to main content

Tutorial: Running a Workflow

Run a small image-processing pipeline, follow its progress, and inspect the outputs. This example uses the bundled Histogram Equalization block to produce enhanced RGB images and a result table.

Before you start​

You need a project with an accessible dataset containing RGB plot assets, an installed histogram_equalization block, and a running worker. Review a few images first. Histogram equalization changes contrast; whether it improves a particular dataset needs visual assessment.

1. Choose the dataset and method​

Open Analysis Modules and find Histogram Equalization. Confirm its version and input requirements. The source example is blocks/image_analysis/histogram_equalization.py; an administrator can install its package if it is missing from the catalog.

Open Analysis Pipelines, select Run pipeline, and use Dataset & Functions to choose the project, study, dataset, and block version.

2. Review the pipeline​

Continue through Functions Ordering and Parameters. This example has one block and does not need a chain of processing steps. In Pipeline, inspect the graph and select Preview JSON to check the dataset and block definition.

When adding more blocks, verify that each output matches the next input. An image saved as an artifact is not necessarily the value returned to the next step.

3. Run and monitor​

Select Run pipeline. Find the new job in the pipeline list and open View output log to follow its progress. The request is queued for the worker, so allow the job to finish before interpreting the outputs.

If it fails, read the error message and confirm the dataset's asset structure, file accessibility, and block dependencies before retrying.

4. Inspect the result​

Open the run's artifacts. The block saves enhanced PNG images and returns a table with one row per processed RGB asset. The worker saves DataFrame results as an Excel workbook. Compare an enhanced image with its original before applying the same workflow more broadly.

Analysis pipelines

Try feature extraction next​

For multispectral data, an installed ndvi_index block can calculate vegetation-index statistics. Check the one-based NIR and red band indices against your sensor's band order. RGB images alone do not supply a near-infrared band.

Continue with Viewing and Exporting Results, or run an analysis conversationally with the Agent.