vsegments 0.1.4 → 0.1.6

package/.claude/settings.local.json

@@ -0,0 +1,12 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(npm rebuild:*)",
+      "Bash(vsegments:*)",
+      "Bash(node -e:*)",
+      "Bash(git add:*)",
+      "Bash(git commit:*)",
+      "Bash(npm publish:*)"
+    ]
+  }
+}

package/bin/cli.js

@@ -11,11 +11,13 @@ const VSegments = require('../src/index');
 
 const program = new Command();
 
+const pkg = require('../package.json');
+
 program
   .name('vsegments')
   .description('Visual segmentation and bounding box detection using Google Gemini AI')
-  .version('0.1.0')
-  .requiredOption('-f, --file <image>', 'Path to input image file')
+  .version(pkg.version)
+  .option('-f, --file <image>', 'Path to input image file')
   .option('--segment', 'Perform segmentation instead of bounding box detection')
   .option('--api-key <key>', 'Google API key (default: GOOGLE_API_KEY env var)')
   .option('-m, --model <model>', 'Model name to use', 'gemini-flash-latest')
@@ -38,6 +40,42 @@ program.parse(process.argv);
 
 const options = program.opts();
 
+// Show welcome message if no file provided
+if (!options.file) {
+  console.log(`
+vsegments v${pkg.version}
+Visual segmentation and bounding box detection using Google Gemini AI
+
+QUICK START
+  vsegments -f image.jpg                   Detect objects with bounding boxes
+  vsegments -f image.jpg --segment         Perform segmentation with masks
+  vsegments -f image.jpg -o output.png     Save visualization to file
+
+CUSTOM PROMPTS
+  vsegments -f photo.jpg -p "find all faces"
+  vsegments -f room.jpg -p "furniture items"
+
+OUTPUT OPTIONS
+  --json results.json    Export detection data as JSON
+  --compact              Print minimal output: "1. label [x y xx yy]"
+  --raw                  Show raw API response
+
+CONFIGURATION
+  --api-key <key>        Google API key (or set GOOGLE_API_KEY env var)
+  --model <name>         Model to use (default: gemini-flash-latest)
+  --temperature <0-1>    Sampling temperature (default: 0.5)
+  --max-objects <n>      Max objects to detect (default: 25)
+
+VISUALIZATION
+  --line-width <n>       Bounding box line width (default: 4)
+  --font-size <n>        Label font size (default: 14)
+  --alpha <0-1>          Mask transparency (default: 0.7)
+
+Run 'vsegments --help' for full options.
+`);
+  process.exit(0);
+}
+
 async function main() {
   try {
     // Validate file exists

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "vsegments",
-  "version": "0.1.4",
+  "version": "0.1.6",
   "description": "Visual segmentation and bounding box detection using Google Gemini AI",
   "main": "src/index.js",
   "types": "src/index.d.ts",
@@ -43,7 +43,8 @@
     "@google/generative-ai": "^0.21.0",
     "canvas": "^2.11.2",
     "commander": "^12.0.0",
-    "sharp": "^0.33.0"
+    "sharp": "^0.33.0",
+    "vsegments": "^0.1.4"
   },
   "devDependencies": {
     "@types/jest": "^30.0.0",

package/src/core.js

@@ -7,11 +7,11 @@ const { loadImage } = require('canvas');
 const fs = require('fs').promises;
 const { SegmentationResult } = require('./models');
 const { parseBoundingBoxes, parseSegmentationMasks } = require('./utils');
-const { 
-  loadImageToCanvas, 
-  plotBoundingBoxes, 
+const {
+  loadImageToCanvas,
+  plotBoundingBoxes,
   plotSegmentationMasks,
-  saveCanvas 
+  saveCanvas
 } = require('./visualize');
 
 class VSegments {
@@ -25,26 +25,38 @@ class VSegments {
    */
   constructor(options = {}) {
     this.apiKey = options.apiKey || process.env.GOOGLE_API_KEY;
-    
+
     if (!this.apiKey) {
       throw new Error(
         'API key must be provided or set in GOOGLE_API_KEY environment variable'
       );
     }
-    
+
     this.model = options.model || 'gemini-3-pro-preview';
     this.temperature = options.temperature !== undefined ? options.temperature : 0.5;
     this.maxObjects = options.maxObjects || 25;
-    
+
     // Initialize Google AI client
     this.genAI = new GoogleGenerativeAI(this.apiKey);
-    
+
     // Default system instructions
     this.defaultSystemInstructions = `
-Return bounding boxes as a JSON array with labels. Never return masks or code fencing. Limit to ${this.maxObjects} objects.
-If an object is present multiple times, name them according to their unique characteristic (colors, size, position, unique characteristics, etc..).
+Return bounding boxes as a JSON array with labels. 
+Never return masks or code fencing. Limit to ${this.maxObjects} objects.
+
+Follow this intuition:
+
+If the object is a face number them according to how someone would draw the features of the face. First the left eye, then the right eye
+then the nose, then the left ear, then the right ear, then the mouth, then the chin.
+Same with Animal faces.
+
+General guideline:
+Follow a drawing order intuitively. People usually do not draw first an eye, and then the background and then the shirt but they
+follow a more natural order with symmetry in mind.
+
+No more than 10 features!
     `.trim();
-    
+
     // Safety settings
     this.safetySettings = [
       {
@@ -53,7 +65,7 @@ If an object is present multiple times, name them according to their unique char
       },
     ];
   }
-  
+
   /**
    * Load image from file and convert to format for API
    * @param {string} imagePath - Path to image file
@@ -62,18 +74,44 @@ If an object is present multiple times, name them according to their unique char
   async _loadImage(imagePath) {
     let imageBuffer = await fs.readFile(imagePath);
     let mimeType = this._getMimeType(imagePath);
-    
+
     // Convert SVG to PNG for API compatibility
     if (mimeType === 'image/svg+xml') {
       const sharp = require('sharp');
-      imageBuffer = await sharp(imageBuffer)
+
+      // Remove common registration/cut line colors from SVG before conversion
+      // These colors are often used for print registration marks, cut lines, etc.
+      let svgString = imageBuffer.toString('utf-8');
+      const registrationColors = [
+        '#ec008c', '#ED008C', // Magenta/pink registration
+        '#00ff00', '#00FF00', // Green registration
+        '#ff0000', '#FF0000', // Red registration (when used for cut lines)
+      ];
+
+      // Remove elements with registration colors
+      for (const color of registrationColors) {
+        // Remove stroke colors
+        const strokeRegex = new RegExp(`stroke="${color}"`, 'gi');
+        svgString = svgString.replace(strokeRegex, 'stroke="none"');
+        // Remove fill colors
+        const fillRegex = new RegExp(`fill="${color}"`, 'gi');
+        svgString = svgString.replace(fillRegex, 'fill="none"');
+      }
+
+      // Increase thin stroke widths for better visibility
+      svgString = svgString.replace(/stroke-width:\s*0\.5pt/gi, 'stroke-width: 2pt');
+      svgString = svgString.replace(/stroke-width="0\.5pt"/gi, 'stroke-width="2pt"');
+
+      imageBuffer = await sharp(Buffer.from(svgString), { density: 300 })
+        .resize(1024, 1024, { fit: 'inside', withoutEnlargement: false })
+        .flatten({ background: { r: 245, g: 245, b: 245 } })  // Light gray background for better contrast
         .png()
         .toBuffer();
       mimeType = 'image/png';
     }
-    
+
     const base64Data = imageBuffer.toString('base64');
-    
+
     return {
       inlineData: {
         data: base64Data,
@@ -81,7 +119,7 @@ If an object is present multiple times, name them according to their unique char
       }
     };
   }
-  
+
   /**
    * Get MIME type from file extension
    * @param {string} filePath - File path
@@ -99,7 +137,7 @@ If an object is present multiple times, name them according to their unique char
     };
     return mimeTypes[ext] || 'image/jpeg';
   }
-  
+
   /**
    * Get system instructions with custom additions
    * @param {string} customInstructions - Additional instructions
@@ -107,14 +145,14 @@ If an object is present multiple times, name them according to their unique char
    */
   _getSystemInstructions(customInstructions) {
     let instructions = this.defaultSystemInstructions;
-    
+
     if (customInstructions) {
       instructions += '\n' + customInstructions;
     }
-    
+
     return instructions;
   }
-  
+
   /**
    * Detect bounding boxes in an image
    * @param {string} imagePath - Path to image file
@@ -130,17 +168,17 @@ If an object is present multiple times, name them according to their unique char
       customInstructions = null,
       maxSize = 1024
     } = options;
-    
+
     // Load image
     const image = await this._loadImage(imagePath);
-    
+
     // Get model
     const model = this.genAI.getGenerativeModel({
       model: this.model,
       safetySettings: this.safetySettings,
       systemInstruction: this._getSystemInstructions(customInstructions)
     });
-    
+
     // Generate content
     let result, response, text;
     try {
@@ -150,7 +188,7 @@ If an object is present multiple times, name them according to their unique char
           temperature: this.temperature,
         }
       });
-      
+
       response = result.response;
       text = response.text();
     } catch (error) {
@@ -161,13 +199,13 @@ If an object is present multiple times, name them according to their unique char
       }
       throw error;
     }
-    
+
     // Parse response
     const boxes = parseBoundingBoxes(text);
-    
+
     return new SegmentationResult(boxes, null, text);
   }
-  
+
   /**
    * Perform segmentation on an image
    * @param {string} imagePath - Path to image file
@@ -181,21 +219,21 @@ If an object is present multiple times, name them according to their unique char
       prompt = 'Give the segmentation masks for the objects. Output a JSON list of segmentation masks where each entry contains the 2D bounding box in the key "box_2d", the segmentation mask in key "mask", and the text label in the key "label". Use descriptive labels.',
       maxSize = 1024
     } = options;
-    
+
     // Load image
     const image = await this._loadImage(imagePath);
-    
+
     // Get image dimensions
     const img = await loadImage(imagePath);
     const imgWidth = img.width;
     const imgHeight = img.height;
-    
+
     // Get model (no system instructions for segmentation)
     const model = this.genAI.getGenerativeModel({
       model: this.model,
       safetySettings: this.safetySettings
     });
-    
+
     // Generate content
     let result, response, text;
     try {
@@ -205,7 +243,7 @@ If an object is present multiple times, name them according to their unique char
           temperature: this.temperature,
         }
       });
-      
+
       response = result.response;
       text = response.text();
     } catch (error) {
@@ -216,14 +254,14 @@ If an object is present multiple times, name them according to their unique char
       }
       throw error;
     }
-    
+
     // Parse response
     const boxes = parseBoundingBoxes(text);
     const masks = await parseSegmentationMasks(text, imgHeight, imgWidth);
-    
+
     return new SegmentationResult(boxes, masks, text);
   }
-  
+
   /**
    * Visualize detection/segmentation results
    * @param {string} imagePath - Path to original image
@@ -242,10 +280,10 @@ If an object is present multiple times, name them according to their unique char
       fontSize = 14,
       alpha = 0.7
     } = options;
-    
+
     // Load image to canvas
     const canvas = await loadImageToCanvas(imagePath, 2048);
-    
+
     // Draw visualizations
     if (result.masks) {
       plotSegmentationMasks(canvas, result.masks, {
@@ -259,12 +297,12 @@ If an object is present multiple times, name them according to their unique char
         fontSize
       });
     }
-    
+
     // Save if requested
     if (outputPath) {
       await saveCanvas(canvas, outputPath);
     }
-    
+
     return canvas;
   }
 }