Azure OCR无法检测到罗马字符“I”和“II” [英] Azure OCR unable to detect roman character “I” and “II”

问题描述

我的图像包含字母I，II，III，IV和V

I have an image which contains letter I, II,III ,IV and V

我正在使用Azure计算机视觉API - v2.0，Azure认知服务网页中提到的识别文本API（POSt）和获取识别文本操作结果（GET）的组合，用于检测图像中的文本字符。

I am using Azure Computer Vision API - v2.0, combination of Recognize Text API(POSt) and Get Recognize Text Operation Result(GET) as mentioned in Azure Cognitive service web page, to detect text characters in the image.

目前它能够检测除字母I和II之外的所有字符。有人可以帮忙吗？

Currently it is able to detect all the characters except letter I and II. Can someone help?

推荐答案

Hello Souvik，

Hello Souvik,

这是一个
图片我用过互联网。 

Here is the result of one image i have used from internet. 

上述文本中的公认文本也在下面提到：

The recognized text from the above text is also mentioned below:

{
  "status": "Succeeded",
  "recognitionResult": {
    "lines": [{
      "boundingBox": [26, 15, 325, 17, 324, 94, 25, 92],
      "text": "Color Sample",
      "words": [{
        "boundingBox": [26, 22, 160, 18, 160, 91, 27, 88],
        "text": "Color",
        "confidence": "Low"
      }, {
        "boundingBox": [156, 18, 330, 18, 329, 95, 156, 91],
        "text": "Sample"
      }]
    }, {
      "boundingBox": [80, 117, 925, 116, 926, 222, 81, 222],
      "text": "I II III IV V",
      "words": [{
        "boundingBox": [63, 123, 127, 122, 126, 221, 62, 221],
        "text": "I"
      }, {
        "boundingBox": [204, 121, 313, 119, 313, 221, 203, 221],
        "text": "II"
      }, {
        "boundingBox": [390, 118, 544, 117, 545, 221, 390, 221],
        "text": "III"
      }, {
        "boundingBox": [615, 117, 736, 117, 738, 222, 615, 221],
        "text": "IV"
      }, {
        "boundingBox": [845, 118, 910, 119, 912, 223, 847, 223],
        "text": "V"
      }]
    }, {
      "boundingBox": [89, 376, 221, 377, 221, 468, 89, 467],
      "text": "VI",
      "words": [{
        "boundingBox": [100, 376, 228, 377, 227, 468, 100, 467],
        "text": "VI"
      }]
    }, {
      "boundingBox": [353, 374, 544, 373, 544, 468, 359, 469],
      "text": "VII",
      "words": [{
        "boundingBox": [377, 372, 550, 371, 551, 467, 378, 468],
        "text": "VII"
      }]
    }, {
      "boundingBox": [688, 376, 914, 373, 914, 469, 689, 470],
      "text": "VIII",
      "words": [{
        "boundingBox": [699, 374, 923, 372, 924, 468, 700, 470],
        "text": "VIII"
      }]
    }, {
      "boundingBox": [87, 627, 221, 631, 219, 718, 90, 716],
      "text": "IX",
      "words": [{
        "boundingBox": [86, 626, 203, 629, 201, 717, 84, 714],
        "text": "IX"
      }]
    }, {
      "boundingBox": [192, 619, 930, 620, 929, 728, 191, 726],
      "text": "X XI XII",
      "words": [{
        "boundingBox": [336, 620, 407, 621, 409, 726, 338, 727],
        "text": "X"
      }, {
        "boundingBox": [536, 622, 673, 624, 675, 723, 538, 724],
        "text": "XI"
      }, {
        "boundingBox": [759, 625, 938, 627, 940, 722, 761, 722],
        "text": "XII"
      }]
    }, {
      "boundingBox": [136, 784, 894, 782, 895, 832, 136, 834],
      "text": "paint your own with Stencils-by-Joanie",
      "words": [{
        "boundingBox": [132, 784, 249, 784, 251, 830, 134, 835],
        "text": "paint"
      }, {
        "boundingBox": [246, 784, 347, 784, 348, 828, 248, 831],
        "text": "your"
      }, {
        "boundingBox": [347, 784, 434, 784, 435, 826, 348, 828],
        "text": "own"
      }, {
        "boundingBox": [440, 784, 527, 784, 528, 826, 442, 826],
        "text": "with"
      }, {
        "boundingBox": [534, 784, 896, 783, 896, 833, 535, 826],
        "text": "Stencils-by-Joanie"
      }]
    }]
  }
}

根据我的经验，我认为不使用标准字体的罗马数字有时可能会被错误识别。

Based on my experience I think the roman numerals which are not using a standard font might sometimes be incorrectly identified.

这篇关于Azure OCR无法检测到罗马字符“I”和“II”的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！