# Vision + OCR: Extract Data from large PDFs with Precision

Hey devs! 👋

Ever needed to pull detailed insights from an image? Whether you’re analyzing documents, tagging photos, or extracting key information, JigsawStack’s [**vOCR API**](https://jigsawstack.com/vocr) is here to make your life easier. With unparalleled accuracy and flexibility, you can now recognize, describe, and retrieve data from images seamlessly.

## What is vOCR?

The vOCR API (Vision Optical Character Recognition) uses advanced AI to analyze images, providing detailed descriptions, metadata, and extracted content. From recognizing text to identifying objects, vOCR opens up endless possibilities for processing visual data.

**Key Features:**

* **Detailed Descriptions:** Generate rich, context-aware narratives for images.
    
* **Flexible Prompts:** Customize what you want to extract from an image.
    
* **High Accuracy:** Reliable recognition across a variety of use cases.
    
* **Scalable API:** Handle large-scale image processing effortlessly.
    

## What Stands Out?

**Contextual Image Analysis**  
vOCR goes beyond basic OCR by analyzing entire images for detailed context not just text.

**Flexible Data Extraction**  
Use custom prompts to retrieve specific details, such as names, objects, or scenes, giving you total control over the output.

## Use Cases

**Document Processing**  
Extract key details from invoices, contracts, or ID cards, such as names and dates.

**Image Tagging**  
Automatically generate tags for large image libraries, perfect for media or e-commerce platforms.

**Scene Understanding**  
Analyze and describe scenes in photos for social media, photography apps, or analytics tools.

**Content Moderation**  
Identify inappropriate or unsafe visual content by analyzing context and objects.

## How to Use the vOCR API

**Step 1:** **Create a free JigsawStack account.**

**Step 2:** **Grab your API key from the dashboard.**

**Step 3: Install the SDK**

Get started by installing the JigsawStack SDK:

```json
npm install jigsawstack
```

Initialize the SDK in your app:

```javascript
import { JigsawStack } from "jigsawstack";

const jigsaw = JigsawStack({
  apiKey: "your-api-key",
});
```

## **Let’s Analyze a PDF**

Extract a detailed description of an image:

```javascript
const result = await jigsaw.vision.vocr({
      url: "https://rogilvkqloanxtvjfrkm.supabase.co/storage/v1/object/public/demo/SG%20firm%20aspires%20to%20be%20Stripe%20of%20AI%20services.pdf?t=2024-11-18T06%3A07%3A00.011Z",
      prompt: "How much time do engineers spend managing APIs?"
    });

console.log("Context:", result.context);
```

**Output**

```bash
Context: Engineers spend up to 80% of their time managing APIs 
instead of building core products.
```

## Let’s Retrieve Specific Data

Use prompts to extract specific details from documents:

```javascript
const result = await jigsaw.vision.vocr({
      url: "https://jigsawstack.com/preview/vocr-example.jpg",
      prompt: ["total_price", "tax", "highlighted_item_name"],  
});
```

**Output**

```bash
Context: {
  total_price: [ '144.02' ],
  tax: [ '4.58' ],
  highlighted_item_name: [ 'GALE' ]
}
```

## Why Choose JigsawStack’s vOCR?

* **Blazing Fast:** Process images with low latency, ideal for real-time applications.
    
* **Flexible and Scalable:** Handle diverse use cases with ease, from general descriptions to specific data extraction.
    
* **Developer-Friendly:** Simple integration with SDKs for python and javascript.
    
* **Secure:** Built-in encryption ensures your data stays safe during processing.
    

With JigsawStack’s vOCR API, you can unlock the full potential of your visual data—whether it’s documents, photos, or complex scenes. Let’s build something amazing together! 🚀