# Find Object - Multiple Image This function block detects whether a template (object) appears across multiple corresponding regions of a larger image. It is designed for cases where both the main image and the template are compared by splitting them into a grid of tiles and checking each tile pair for a match. The block provides visual feedback by marking tiles that failed to match and outputs both counts and an overall pass/fail indicator. ## đŸ“Ĩ Inputs `Main Image` Provide the image in which you want to verify the presence of the object.\ `Object Image` Provide the template image that should be present across the main image tiles. ## 📤 Outputs `Image Any` The main image annotated to indicate tiles where the template was not found (visual feedback).\ `Match Count` The number of tiles where the template was successfully matched.\ `Matched All` Boolean that is TRUE when the template is found in every expected tile (based on the chosen slice grid), otherwise FALSE. ## 🕹ī¸ Controls `Method` Choose the matching method to control how image similarity is evaluated between each tile pair. Different methods change matching sensitivity and behavior. `Match Threshold %` Adjust how strict the matching is. Increasing the threshold requires closer similarity for a tile to count as matched. `Horizontal Slice` Enter how many tiles the image should be split horizontally (columns). `Vertical Slice` Enter how many tiles the image should be split vertically (rows). ## đŸŽ¯ Key Features * Grid-based comparison: Splits both main and template images into the same number of tiles and compares corresponding tiles. * Visual failure indicators: The output image is annotated to show tiles that did not match (easy visual inspection). * Match summary: Returns numeric count of matched tiles and a boolean `Matched All` to quickly know if the entire grid passed. * Adjustable sensitivity and method: Fine-tune the detection behavior using the `Method` and `Match Threshold %` controls. * Works with multiple scales via tile-based comparison, useful when the object should appear repeatedly across a structured layout. ## ⚙ī¸ How it runs When the block runs it: 1. Splits both the `Main Image` and the `Object Image` into a grid defined by `Horizontal Slice` and `Vertical Slice`. 2. Compares each pair of corresponding tiles using the selected `Method` and the `Match Threshold %`. 3. Marks tiles on the output image where the template was not found and increments a match counter for successful tiles. 4. Outputs the annotated image, the total matched tile count, and a boolean indicating whether all tiles matched. Note: For reliable results the main and object images should be similarly scaled and oriented so that corresponding tiles are comparable. ## 📝 Usage Tips * Choose `Horizontal Slice` and `Vertical Slice` to match the expected tiling or layout of the object in the main image (for example, a 3x3 grid for nine expected positions). * Increase `Match Threshold %` to reduce false positives when background or pattern noise is present. * Try different `Method` options if one method produces too many misses or false matches. ## 💡 Tips and Tricks — Useful block combinations * If your images are too large or slow to process, use `Image Resizer` before this block to reduce input size. * Use `Image ROI Select` to crop the main image to the area of interest so slicing only applies where needed. * Apply `Blur` or `Denoising` to the main image to reduce high-frequency noise that may cause false mismatches. * Adjust colors or contrast with `Adjust Colors` or `Contrast Optimization` to make the object more distinguishable from the background. * Preview results by sending the annotated output to `Show Image`. * Save failing cases or the annotated image using `Image Logger` or `Image Write` for offline review. * For single-template checks (no tiling), consider `Find Object` instead when you only need to check for one occurrence rather than a tiled grid. (hint: combine the above controls and blocks to build a robust inspection flow — crop, preprocess, compare, then log or display results) ## 🛠ī¸ Troubleshooting * If the block reports fewer matches than expected: verify `Horizontal Slice` and `Vertical Slice` match how the object repeats in the main image; increase `Match Threshold %` or try a different `Method`. * If the block reports too many false matches: raise the `Match Threshold %`, apply preprocessing such as `Image Threshold` or `Blur`, or use `Image Resizer` to align scales. * If slice counts do not match or comparison seems invalid: ensure both inputs are valid images and the slice numbers are positive integers. * If orientation or scale differs between template and main image: rotate or resize inputs so tiles align, or use `Auto Alignment` / `Image Resizer` before running comparison. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.augelab.com/function-blocks/detections-shapes/detectors/find-object-multiple-image.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.