无法获得Vision API v1.0 Describe Image以使用Raw图像二进制文件 [英] Unable to get Vision API v1.0 Describe Image to work with Raw image binary
问题描述
我正在尝试使用认知服务愿景1.0'描述'功能，但我不能除了传递< g class =" gr_ gr_34 gr-alert gr_spell gr_inline_cards gr_run_anim ContextualSpelling ins-del multiReplace"
data-gr-id =" 34" ID = QUOT; 34"> URL< / g取代;在邮政体内。文档说你应该能够传递'原始图像二进制文件'，但是我在使用该选项时没有成功。
我已经取得了成功可以通过搜索azure认知服务找到python示例（显然我不能使用链接....）计算机视觉快速启动python-disk
我可以创建看似正确的'原始图像二进制'，但我不能在这一行创建字节数组：
＃将图像读入一个字节
arrayimage_data = open（image_path，" rb"）。read（）
$
$
并使用API工具或邮递员创建相同的响应...&
$
有人能提供一个"原始图像二进制"字符串的示例，该字符串在python示例之外成功运行吗？ br $>
$
这就是我要找的东西：
POST<我不能使用链接，但这是端点>
主机：<我不能使用链接，但这是主机>
内容类型：application /< g class =" gr_ gr_35 gr-alert gr_spell gr_inline_cards gr_run_anim ContextualSpelling"数据-GR-ID =" 35 QUOT; id =" 35"> json< / g>
{" data"："< Put Raw Binary Here>"}
应该产生：
{
&NBSP; "description"：{
&NBSP; &NBSP; &NBSP; "tags"：[
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "室内"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "table"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "坐着"，
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "hold"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "笔记本电脑"，
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "man"，
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "computer"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "使用"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "女人"，
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "living"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "black"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "white"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "年轻"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "房间"，
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "desk"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "play"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "board"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "people"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "video"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "remote"&b
&NBSP; &NBSP; &NBSP; ]，
&NBSP; &NBSP; &NBSP; "字幕"：[
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; {
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "text"："坐在桌子上的人"，"
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; &NBSP; &NBSP; "置信度"：0.41177881844649067
&NBSP; &NBSP; &NBSP; &NBSP; &NBSP; }¥b $ b &NBSP; &NBSP; &NBSP; ]¥b $ b &NBSP; }，b $ b &NBSP; "requestId"：" 45e02d4a-f983-4222-a102-57d4c6e94158"，&
&NBSP; "元数据"：{
&NBSP; &NBSP; &NBSP; "宽度"：2400，
&NBSP; &NBSP; &NBSP; "身高"：1263，
&NBSP; &NBSP; &NBSP; "格式"：" Jpeg"
&NBSP; }
}
$
这是其他人试图做同样的事情，但放弃了：
https://github.com/Microsoft/BotBuilder/issues/3188#issuecomment-318970877
I'm trying to use the Cognitive Services vision 1.0 'Describe' functionality, but I'm not able to get it to work with anything except passing a <g class="gr_ gr_34 gr-alert gr_spell gr_inline_cards gr_run_anim ContextualSpelling ins-del multiReplace"
data-gr-id="34" id="34">url</g> in the post body. The documentation says you should be able to pass a 'raw image binary' but I've had zero success in using that option.
I have had success with the python example (apparently I can't use links yet....) which can be found by searching for azure cognitive-services Computer-vision quickstarts python-disk
I can create what seems to be the correct 'raw image binary', but I can't take the byte array created on this line:
# Read the image into a byte
arrayimage_data = open(image_path, "rb").read()
And use the API tool or Postman to create the same response...
Can anyone provide an example of a 'Raw image binary' string that works successfully outside of the python example?
This is kind of what I'm looking for:
POST <I can't use a link, but this is the endpoint>
Host: <I can't use a link, but this is the host>
Content-Type: application/<g class="gr_ gr_35 gr-alert gr_spell gr_inline_cards gr_run_anim ContextualSpelling" data-gr-id="35" id="35">json</g>
{"data":"<Put Raw Binary Here>"}
Should produce:
{
"description": {
"tags": [
"indoor",
"table",
"sitting",
"holding",
"laptop",
"man",
"computer",
"using",
"woman",
"living",
"black",
"white",
"young",
"room",
"desk",
"playing",
"board",
"people",
"video",
"remote"
],
"captions": [
{
"text": "a person sitting on a table",
"confidence": 0.41177881844649067
}
]
},
"requestId": "45e02d4a-f983-4222-a102-57d4c6e94158",
"metadata": {
"width": 2400,
"height": 1263,
"format": "Jpeg"
}
}
Here is someone else trying to do the same thing, but gave up:
https://github.com/Microsoft/BotBuilder/issues/3188#issuecomment-318970877
这篇关于无法获得Vision API v1.0 Describe Image以使用Raw图像二进制文件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！
