planeta ai logo
Darmowy szablon automatyzacji

Porównaj lokalne modele widzenia Ollama do analizy obrazu za pomocą Dokumentów Google

1514
27 dni temu
19
bloków


Compare Local Ollama Vision Models for Image Analysis using Google Docs

Przetwarzaj obrazy przy użyciu lokalnie hostowanych modeli Ollama Vision, aby wyodrębnić szczegółowe opisy, kontekstowe informacje i ustrukturyzowane dane. Wyniki zapisuj bezpośrednio w Google Docs, co ułatwia współpracę.

Dla kogo jest ten workflow?

Ten workflow jest idealny dla programistów, analityków danych, marketerów i entuzjastów sztucznej inteligencji, którzy potrzebują przetwarzać i analizować obrazy przy użyciu lokalnie hostowanych modeli Ollama Vision Language Models. Szczególnie przydatny jest w zadaniach wymagających szczegółowych opisów obrazów, analizy kontekstowej i ekstrakcji ustrukturyzowanych danych.

Jaki problem rozwiązuje ten workflow?

Workflow rozwiązuje wyzwanie związane z wyodrębnianiem znaczących informacji z obrazów w szczegółowy sposób, takich jak identyfikacja obiektów, analiza relacji przestrzennych, ekstrakcja elementów tekstowych i dostarczanie informacji kontekstowych. Jest to szczególnie pomocne w zastosowaniach w nieruchomościach, marketingu, inżynierii i badaniach.

Co robi ten workflow?

  • Pobiera plik obrazu z Google Drive.
  • Przetwarza obraz przy użyciu wielu modeli Ollama Vision (np. Granite3.2-Vision, Gemma3, Llama3.2-Vision).
  • Generuje szczegółowe opisy obrazów w formacie markdown.
  • Zapisuje wyniki do pliku Google Docs w celu łatwego udostępniania i dalszej analizy.

Konfiguracja

  • Upewnij się, że masz dostęp do lokalnej instancji Ollama: https://ollama.com/.
  • Pobierz modele Ollama Vision.
  • Skonfiguruj swoje dane uwierzytelniające Google Drive i Google Docs w n8n.
  • Podaj ID pliku obrazu z Google Drive w odpowiednim węźle.
  • Zaktualizuj listę modeli Ollama Vision.
  • Przetestuj workflow, klikając "Test Workflow", aby uruchomić proces.

Jak dostosować ten workflow do swoich potrzeb?

  • Zastąp źródło obrazu innym dostawcą, jeśli to konieczne (np. AWS S3 lub Dropbox).
  • Zmodyfikuj prompty w węźle "General Image Prompt", aby dostosować je do konkretnych wymagań analizy.
  • Dodaj dodatkowe węzły do przetwarzania końcowego lub integracji wyników z innymi platformami, takimi jak Slack czy HubSpot.

Kluczowe funkcje

  • Szczegółowa analiza obrazów: Wyodrębnia kompleksowe informacje o obiektach, relacjach przestrzennych, elementach tekstowych i ustawieniach kontekstowych.
  • Obsługa wielu modeli: Wykorzystuje dynamicznie wiele modeli wizyjnych dla optymalnej wydajności.
  • Wyniki w formacie markdown: Formatuje wyniki w markdown dla łatwej czytelności i dokumentacji.
  • Integracja z Google Drive: Bezproblemowo pobiera obrazy i zapisuje wyniki w Google Docs.

Przykłady zastosowań

Ten workflow może być wykorzystany w wielu różnych scenariuszach, w tym:

  • Nieruchomości: Analiza zdjęć nieruchomości w celu identyfikacji kluczowych cech i uszkodzeń.
  • Marketing: Automatyczne generowanie opisów produktów na podstawie zdjęć.
  • Inżynieria: Wykrywanie wad lub nieprawidłowości na zdjęciach technicznych.
  • Badania naukowe: Analiza obrazów mikroskopowych lub satelitarnych.
  • E-commerce: Tworzenie szczegółowych opisów produktów na podstawie zdjęć.
  • Bezpieczeństwo: Monitorowanie i analiza obrazów z kamer bezpieczeństwa.
  • Edukacja: Automatyczne generowanie opisów dla materiałów edukacyjnych.


   Skopiuj kod szablonu   
{"id":"keFEBUqHOrsib60G","meta":{"instanceId":"31e69f7f4a77bf465b805824e303232f0227212ae922d12133a0f96ffeab4fef","templateCredsSetupCompleted":true},"name":"🦙👁️👁️ Find the Best Local Ollama Vision Models by Comparison","tags":[],"nodes":[{"id":"dd2f1201-a78a-4ea9-b5ff-7543673e8445","name":"Sticky Note1","type":"n8n-nodes-base.stickyNote","position":[1080,1160],"parameters":{"color":4,"width":340,"height":340,"content":"## 👁️ Analyze Image with Local Ollama LLMn"},"typeVersion":1},{"id":"81975be4-1e40-41e9-b938-612270f80a92","name":"Sticky Note2","type":"n8n-nodes-base.stickyNote","position":[280,640],"parameters":{"color":4,"width":300,"height":300,"content":"## 👍Try Me!"},"typeVersion":1},{"id":"3a56f75b-4836-4c37-a246-83ef6507c581","name":"When clicking ‘Test workflow’","type":"n8n-nodes-base.manualTrigger","position":[380,740],"parameters":{},"typeVersion":1},{"id":"bb4570c7-269c-4d28-85d4-183ca2fabb89","name":"Ollama LLM Request","type":"n8n-nodes-base.httpRequest","position":[1200,1280],"parameters":{"url":"http://127.0.0.1:11434/api/chat","method":"POST","options":{},"jsonBody":"={{ $json.body }}","sendBody":true,"sendHeaders":true,"specifyBody":"json","headerParameters":{"parameters":[{"name":"Content-Type","value":" application/json"}]}},"typeVersion":4.2},{"id":"0a6e064d-9a67-4cd2-b3ed-247a1684c1fb","name":"Create Request Body","type":"n8n-nodes-base.set","position":[840,1280],"parameters":{"options":{},"assignments":{"assignments":[{"id":"be9a8e21-9bb6-4588-a77a-61bc2def0648","name":"body","type":"string","value":"={n "model": "{{ $json.models }}",n "messages": [n {n "role": "user",n "content": "{{ $json.user_prompt }}",n "images": ["{{ $('List of Vision Models').item.json.data }}"]n }n ],n "stream": falsen}"}]}},"typeVersion":3.4},{"id":"f3119aff-62bc-4ffc-abe9-835dea105d76","name":"Loop Over Ollama Models","type":"n8n-nodes-base.splitInBatches","position":[360,1180],"parameters":{"options":{}},"typeVersion":3},{"id":"1e26f493-881e-40a3-922d-5c8d6cb86374","name":"Create Result Objects","type":"n8n-nodes-base.set","position":[620,1080],"parameters":{"options":{},"assignments":{"assignments":[{"id":"780086e5-2733-435a-90b5-fd10946ddd7a","name":"result","type":"object","value":"={{ $json }}"}]}},"typeVersion":3.4},{"id":"ac0d3ada-8890-4945-aedb-fd6be4ffc020","name":"General Image Prompt","type":"n8n-nodes-base.set","position":[620,1280],"parameters":{"options":{},"assignments":{"assignments":[{"id":"be9a8e21-9bb6-4588-a77a-61bc2def0648","name":"user_prompt","type":"string","value":"=Analyze this image in exhaustive detail using this structure:\n\n1. **Comprehensive Inventory**\n- List all visible objects with descriptors (size, color, position)\n- Group related items hierarchically (primary subject → secondary elements → background)\n- Note object conditions (intact/broken, new/aged)\n\n2. **Contextual Analysis**\n- Identify probable setting/location with supporting evidence\n- Determine time period/season through visual cues\n- Analyze lighting conditions and shadow orientation\n\n3. **Spatial Relationships**\n- Map object positions using grid coordinates (front/center/back, left/right)\n- Describe size comparisons between elements\n- Note overlapping/occluded objects\n\n4. **Textual Elements**\n- Extract ALL text with font characteristics\n- Identify logos/brands with confidence estimates\n- Translate non-native text with cultural context\n\nFormat response in markdown with clear section headers and bullet points."}]},"includeOtherFields":true},"typeVersion":3.4},{"id":"b7faafca-a179-43b2-8318-29b4659d424f","name":"Real Estate Spreadsheet Prompt","type":"n8n-nodes-base.set","disabled":true,"position":[620,1480],"parameters":{"options":{},"assignments":{"assignments":[{"id":"be9a8e21-9bb6-4588-a77a-61bc2def0648","name":"user_prompt","type":"string","value":"=Analyze this spreadsheet image in exhaustive detail using this structure:\n\n1. **Table Structure**\n- Identify all column headers (months) in order\n- List all row labels exactly as shown\n- Note any table titles, footnotes, or metadata\n\n2. **Data Extraction**\n- Extract all numeric values with precise formatting (decimals, currency symbols)\n- Maintain exact numbers for Listings, Sales, Months of Inventory\n- Preserve currency formatting for Avg. Price values\n- Include DOM values from separate section\n\n3. **Markdown Representation**\n- Convert the entire spreadsheet into a perfectly formatted markdown table\n- Maintain alignment of all columns and rows\n- Preserve all relationships between data points\n\n4. **Data Analysis**\n- Identify trends across months for each metric\n- Note highest and lowest values in each category\n- Calculate percentage changes between months where relevant\n\nFormat response with a complete markdown table first, followed by brief analysis of the real estate market data shown."}]},"includeOtherFields":true},"typeVersion":3.4},{"id":"5c28053b-a44e-494b-ad03-27d5b217f6b3","name":"List of Vision Models","type":"n8n-nodes-base.set","position":[1440,740],"parameters":{"options":{},"assignments":{"assignments":[{"id":"86add667-cd96-4e1c-877a-c437f6b1e040","name":"models","type":"array","value":"=["granite3.2-vision","llama3.2-vision","gemma3:27b"]"}]},"includeOtherFields":true},"typeVersion":3.4},{"id":"58c2fcd2-0ac4-4684-a30f-37650cc8dac1","name":"Get Base64 String","type":"n8n-nodes-base.extractFromFile","position":[1140,740],"parameters":{"options":{},"operation":"binaryToPropery"},"typeVersion":1},{"id":"b60c0589-397c-445b-a084-a791bef95b15","name":"Download Image File from Google Drive","type":"n8n-nodes-base.googleDrive","position":[920,740],"parameters":{"fileId":{"__rl":true,"mode":"id","value":"={{ $json.id }}"},"options":{},"operation":"download"},"credentials":{"googleDriveOAuth2Api":{"id":"UhdXGYLTAJbsa0xX","name":"Google Drive account"}},"typeVersion":3},{"id":"55a8f511-fdb5-4830-837a-104cbf6c6167","name":"Split List of Vision Models for Looping","type":"n8n-nodes-base.splitOut","position":[1640,740],"parameters":{"options":{},"fieldToSplitOut":"models"},"typeVersion":1},{"id":"8e48c8bd-15c9-4389-8698-77dc5ae698bc","name":"Sticky Note","type":"n8n-nodes-base.stickyNote","position":[620,640],"parameters":{"color":7,"width":700,"height":300,"content":"## ⬇️Download Image from Google Drive"},"typeVersion":1},{"id":"6ed8925d-b031-4052-9009-91e2e7d8f360","name":"Sticky Note3","type":"n8n-nodes-base.stickyNote","position":[1360,640],"parameters":{"color":7,"width":460,"height":300,"content":"## 📜Create List of Local Ollama Vision Models"},"typeVersion":1},{"id":"ae383e4f-21e6-479f-97e0-029f43dacc56","name":"Sticky Note4","type":"n8n-nodes-base.stickyNote","position":[280,980],"parameters":{"color":7,"width":1200,"height":720,"content":"## 🦙👁️👁️ Process Image with Ollama Vision Models and Save Results to Google Drive"},"typeVersion":1},{"id":"a27bcb6e-c6e8-4777-9887-428363256b4a","name":"Google Doc Image Id","type":"n8n-nodes-base.set","position":[700,740],"parameters":{"options":{},"assignments":{"assignments":[{"id":"7d5a0385-4d8b-4f70-b3b0-4182bda29e5c","name":"id","type":"string","value":"=[your-google-id]"}]}},"typeVersion":3.4},{"id":"8e6114f8-c724-40fd-9be3-253e3cb882fa","name":"Save Image Descriptions to Google Docs","type":"n8n-nodes-base.googleDocs","position":[840,1080],"parameters":{"actionsUi":{"actionFields":[{"text":"=<{{ $json.result.model }}>n{{ $json.result.message.content }}nnn","action":"insert"}]},"operation":"update","documentURL":"[your-google-doc-id]"},"credentials":{"googleDocsOAuth2Api":{"id":"YWEHuG28zOt532MQ","name":"Google Docs account"}},"typeVersion":2},{"id":"abed9af8-0d50-413a-9e6d-c6100ddaf015","name":"Sticky Note5","type":"n8n-nodes-base.stickyNote","position":[-240,640],"parameters":{"width":480,"height":1340,"content":"## 🦙👁️👁️ Find the Best Local Ollama Vision Models for Your Use CasennProcess images using locally hosted Ollama Vision Models to extract detailed descriptions, contextual insights, and structured data. Save results directly to Google Docs for efficient collaboration.nn### Who is this for?nThis workflow is ideal for developers, data analysts, and AI enthusiasts who need to process and analyze images using locally hosted Ollama Vision Language Models. It’s particularly useful for tasks requiring detailed image descriptions, contextual analysis, and structured data extraction.nn### What problem is this workflow solving? / Use CasenThe workflow solves the challenge of extracting meaningful insights from images in exhaustive detail, such as identifying objects, analyzing spatial relationships, extracting textual elements, and providing contextual information. This is especially helpful for applications in real estate, marketing, engineering, and research.nn### What this workflow doesnThis workflow:n1. Downloads an image file from Google Drive.n2. Processes the image using multiple Ollama Vision Models (e.g., Granite3.2-Vision, Llama3.2-Vision).n3. Generates detailed markdown-based descriptions of the image.n4. Saves the output to a Google Docs file for easy sharing and further analysis.nn### Setupn1. Ensure you have access to a local instance of Ollama. https://ollama.com/n2. Pull the Ollama vision models.n3. Configure your Google Drive and Google Docs credentials in n8n.n4. Provide the image file ID from Google Drive in the designated node.n5. Update the list of Ollama vision modelsn6. Test the workflow by clicking ‘Test Workflow’ to trigger the process.nn### How to customize this workflow to your needsn- Replace the image source with another provider if needed (e.g., AWS S3 or Dropbox).n- Modify the prompts in the "General Image Prompt" node to suit specific analysis requirements.n- Add additional nodes for post-processing or integrating results into other platforms like Slack or HubSpot.nn## Key Features:n- **Detailed Image Analysis**: Extracts comprehensive details about objects, spatial relationships, text elements, and contextual settings.n- **Multi-Model Support**: Utilizes multiple vision models dynamically for optimal performance.n- **Markdown Output**: Formats results in markdown for easy readability and documentation.n- **Google Drive Integration**: Seamlessly downloads images and saves results to Google Docs.nnn"},"typeVersion":1}],"active":false,"pinData":{},"settings":{"executionOrder":"v1"},"versionId":"a337e019-1c9a-4736-8dcd-4f12a9d989f4","connections":{"Get Base64 String":{"main":[[{"node":"List of Vision Models","type":"main","index":0}]]},"Ollama LLM Request":{"main":[[{"node":"Loop Over Ollama Models","type":"main","index":0}]]},"Create Request Body":{"main":[[{"node":"Ollama LLM Request","type":"main","index":0}]]},"Google Doc Image Id":{"main":[[{"node":"Download Image File from Google Drive","type":"main","index":0}]]},"General Image Prompt":{"main":[[{"node":"Create Request Body","type":"main","index":0}]]},"Create Result Objects":{"main":[[{"node":"Save Image Descriptions to Google Docs","type":"main","index":0}]]},"List of Vision Models":{"main":[[{"node":"Split List of Vision Models for Looping","type":"main","index":0}]]},"Loop Over Ollama Models":{"main":[[{"node":"Create Result Objects","type":"main","index":0}],[{"node":"General Image Prompt","type":"main","index":0}]]},"Real Estate Spreadsheet Prompt":{"main":[[]]},"When clicking ‘Test workflow’":{"main":[[{"node":"Google Doc Image Id","type":"main","index":0}]]},"Download Image File from Google Drive":{"main":[[{"node":"Get Base64 String","type":"main","index":0}]]},"Split List of Vision Models for Looping":{"main":[[{"node":"Loop Over Ollama Models","type":"main","index":0}]]}}}
  • API
  • Request
  • URL
  • Build
  • cURL
Planeta AI 2025 
magic-wandmenu linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram