> For the complete documentation index, see [llms.txt](https://docs.augelab.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.augelab.com/function-blocks/blocks-reference/image-transformations/analysis/histogram-on-curve.md). # Histogram On Curve This function block analyzes pixel values along a detected curved line across a grayscale or binary image. It detects line points, fits a smooth curve, evaluates pixel values along that curve against a threshold, and returns visualizations and numeric results to help you inspect line quality and detect gaps or peaks. ## đŸ“Ĩ Inputs `Input Image (Binary Image)` Grayscale or binary image to analyze. Preferred when foreground/background contrast is strong. `Line Location (pixel)` Number of sample lines to use when detecting the curve (the block samples multiple straight lines and derives curve points). `Line detection threshold` Pixel intensity threshold used during line pixel detection (valid range 0–255). `Pixel Value Threshold` Pixel intensity threshold used when evaluating pixel values along the detected curve (valid range 0–255). ## 📤 Outputs `Output Image` Annotated source image showing the sampled lines/curve and detected points. `Plot Image` Plot visualization showing pixel profiles and gradient information used during detection. `Curve Image` Visualization of the detected curve and pixel values along it. `Line Angle` Estimated angle of the fitted curve (useful for geometry checks). `Is Line Valid?` Boolean flag indicating whether the detected curve passes internal validity checks (useful to gate further processing). ## 🕹ī¸ Controls `Relation` Dropdown to select line sampling relation (vertical vs horizontal orientation). This changes how the block samples image lines during curve detection. ## ✨ Features * Detects curve points by sampling several straight lines across the image and fitting a smooth polynomial through these points. * Produces both image annotations and plotted diagnostics to help verify detection quality. * Validity check that flags unreliable detections so downstream blocks can handle failures. * Adjustable thresholds and line sampling count via inputs for fine control over sensitivity. ## 📝 Usage Instructions 1. Provide a clear grayscale or binary image to `Input Image (Binary Image)`. 2. Set `Line Location (pixel)` to the number of sampling lines (higher numbers increase robustness but are slower). 3. Tune `Line detection threshold` to reliably identify line pixels on sampled lines. 4. Tune `Pixel Value Threshold` to mark significant pixel regions along the fitted curve. 5. Inspect `Output Image`, `Plot Image`, and `Curve Image` to validate detection. Use `Is Line Valid?` to decide whether to continue processing. ## 📊 Evaluation When evaluated, the block: * Samples multiple straight lines across the image (orientation chosen via `Relation`), * Detects line pixels and computes curve points, * Fits a smooth curve and computes angle and pixel statistics, * Returns annotated images, plots, and validation flags for downstream use. ## 💡 Tips and Tricks * Preprocess noisy images with `Blur` to reduce spurious peaks before curve detection. * Use `Image Threshold` to create a binary image if the curve contrast is low — this often improves line detection reliability. * Crop the area of interest with `Image ROI Select` so the block focuses on the relevant region and avoids unrelated edges. * Send visual outputs to `Show Image` to inspect `Output Image` / `Plot Image` / `Curve Image` interactively while tuning parameters. * If you want to save analysis results for later review, use `Image Logger` or `Image Write` with the images returned from this block. (hint: combine preprocessing blocks such as `Blur` and `Image Threshold` upstream, then feed into this block; visualize results with `Show Image` and optionally archive them with `Image Logger` or `Image Write`.) ## 🛠ī¸ Troubleshooting * If you get an error about invalid line detection, try increasing the number of sampled lines (`Line Location (pixel)`) or preprocess with `Blur` and `Image Threshold`. * If no peaks appear in plots, reduce thresholds or improve contrast of the `Input Image (Binary Image)`. * If the detected curve is shifted or noisy, crop the image with `Image ROI Select` to remove distracting regions and try again.