Agricultural Crop Anomaly Detection Tool

Name: Agricultural Crop Anomaly Detection Tool
Rating: 5 (5 reviews)
Author: Free N8N

Beginner

8 nodes connected

detail.loadingPreview

Free N8N Temples

81 views

0 downloads

AI & Machine Learningagricultureanomaly-detectionimage-recognitionqdrantvoyageai

Detects if an input image represents an anomalous crop not present in a pre-defined dataset.

About This Workflow

This workflow acts as an anomaly detection tool for agricultural crops. It takes an image URL as input and leverages embeddings and a vector database (Qdrant) to determine if the depicted crop is anomalous compared to a known dataset of agricultural crops. The system compares the input image's embedding against pre-calculated cluster medoids and their associated threshold scores within the Qdrant collection. If the image's similarity score falls below the established thresholds for all known crop classes, it is flagged as an anomaly.

Key Features

Image Anomaly Detection: Identifies crops not belonging to a predefined set.
Multimodal Embeddings: Uses Voyage AI's multimodal embeddings API to generate vector representations of images.
Vector Database Integration: Queries a Qdrant collection to find similar crop embeddings.
Threshold-Based Classification: Compares similarity scores against pre-set thresholds for each crop class.
Configurable Variables: Allows setting Qdrant cluster types and API endpoints.

How To Use

Trigger: The workflow is initiated via the Execute Workflow Trigger node. Provide the imageURL of the image you want to analyze in the pinData section.
Image Embedding: The Image URL hardcode node prepares the image URL, and the Embed image node sends it to the Voyage AI API to generate a multimodal embedding.
Qdrant Query: The Variables for medoids node sets up necessary variables for Qdrant. The Total Points in Collection and Each Crop Counts nodes are used to determine the number of crops, which is then passed to Info About Crop Labeled Clusters as cropsNumber.
Similarity Check: The Get similarity of medoids node queries the Qdrant collection using the generated embedding. It searches for points that match the specified clusterCenterType and retrieves a number of results equal to cropsNumber.
Anomaly Determination: The Compare scores node analyzes the similarity scores from Qdrant. It checks if the score for any cluster exceeds its respective threshold (clusterThresholdCenterType). If no cluster's score meets or exceeds its threshold, an alert is generated indicating a potential new undefined crop. Otherwise, it suggests the most similar crop class.

Apps Used

agriculture

anomaly-detection

image-recognition

qdrant

voyageai

Workflow JSON

{
  "id": "e57af55c-7599-4bb4-93e2-1238950d2876",
  "name": "Agricultural Crop Anomaly Detection Tool",
  "nodes": 8,
  "category": "AI & Machine Learning",
  "status": "active",
  "version": "1.0.0"
}

Note: This is a sample preview. The full workflow JSON contains node configurations, credentials placeholders, and execution logic.