Environment details

• OS: MacOS
• Node.js version: 8.1.4
• npm version: 5.0.3
• @google-cloud/vision version: 0.11.4

Steps to reproduce

1. download the sample image
   

As it is pretty clear, you'd expect it to return the capital letter "M" in the text detection field.

2. Fire up a text editor and run the below node snippet in a env that already has GCloud authentification setup. You will also need the Vision API enabled

'use strict';

const vision = require('@google-cloud/vision')();

const options = {
  imagePath: __dirname + '/sample.png'
}

vision.detectText(options.imagePath).then((result) => {
  console.log('this is the result', JSON.stringify(result));
});

Now the result looks like this:

[
   [

   ],
   {
      "responses":[
         {
            "faceAnnotations":[],
            "landmarkAnnotations":[],
            "logoAnnotations":[],
            "labelAnnotations":[],
            "textAnnotations":[],
            "fullTextAnnotation":null,
            "safeSearchAnnotation":null,
            "imagePropertiesAnnotation":null,
            "cropHintsAnnotation":null,
            "webDetection":null,
            "error":null
         }
      ]
   }
]

If i test the same sample image on GCloud Vision landing page demo, the result is correct: the letter M

https://cloud.google.com/vision/

Since this is a pretty clear image and 1 single character, I'm having trouble understanding why does the accuracy differ.
I noticed this issue also on other sample images - they work in the demo but not using the Node SDK.