Computer Vision

Welcome to the Computer Vision API Forum

Categories

API – Any ideas or feedback pertaining to features or enhancements to Computer Vision API.

Documentation – Any ideas or suggestions for the API Reference or Documentation.

Language Support – Submit a request to have a particular language supported.

Samples & SDK Request – Let us know if you would like to see a Code sample or SDK provided.

Custom/Sample Images – Have an image you’ve tested and not getting the results you are seeking? Upload the image and describe the information or tags you would like to be included.

How can we improve Computer Vision?

(thinking…)

Enter your idea and we'll search to see if someone has already suggested it.

If a similar idea already exists, you can support and comment on it.

If it doesn't exist, you can post your idea so others can support it.

Enter your idea and we'll search to see if someone has already suggested it.

  1. Pattern definition

    We have some standard forms that are in handwritten format but the pattern of the text is always the same (I think this can apply to a large number of scenarios). Allowing the handwritten recognizer to be configured with a expected pattern via a simple regex could improve a lot of scenarios.

    2 votes
    Sign in
    Check!
    (thinking…)
    Reset
    or sign in with
    • facebook
    • google
      Password icon
      I agree to the terms of service
      Signed in as (Sign out)

      We’ll send you updates on this idea

      0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
    • How could I use http post enquery of C languege to get the Computer Vision API ?

      postRequest =(String)("POST ") +"/vision/v1.0/analyze?https://upload.wikimedia.org/wikipedia/commons/1/12/Broadway_and_Times_Square_by_night.jpg HTTP/1.1\n"
      + "visualFeatures: Categories,Description,Color"
      + "language: en\n"
      + "Content-Type: application/json"
      + "Ocp-Apim-Subscription-Key: {mysub key}\n"
      + "Host: westus.api.cognitive.microsoft.com\n"
      + "Content-Length: 152\r\n\r\n";

      As shown above, I'll get an error number of 401,which means Access denied due to missing subscription key , how do I organize the content of my post request ?

      1 vote
      Sign in
      Check!
      (thinking…)
      Reset
      or sign in with
      • facebook
      • google
        Password icon
        I agree to the terms of service
        Signed in as (Sign out)

        We’ll send you updates on this idea

        0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
      • Offer relative position in 3D-space between two or more images

        If I take two photos of a subject, I would love to be able to know where the second photo was taken in real space relative to the first. In the example of handheld smartphone photography, if the first image is a baseline then I would love to know the xyz offset of the second image. Even if I believe that I'm holding a smartphone in the same position without moving it, it's likely moving at least an inch in multiple dimensions just due to hand shake. However, I might also take a side profile and it would be just…

        1 vote
        Sign in
        Check!
        (thinking…)
        Reset
        or sign in with
        • facebook
        • google
          Password icon
          I agree to the terms of service
          Signed in as (Sign out)

          We’ll send you updates on this idea

          0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
        • Increase face landmark density

          Currently using Dlib facial detection which returns 68 facial landmarks vs Azure's 27. More is better; this is proving to be a political issue with the system architect.

          1 vote
          Sign in
          Check!
          (thinking…)
          Reset
          or sign in with
          • facebook
          • google
            Password icon
            I agree to the terms of service
            Signed in as (Sign out)

            We’ll send you updates on this idea

            0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
          • Offer subject foreground/background segmentation

            Being able to reliably pull out complex bounding areas would be a game changer, ideally if the classifier had an understanding of human proportion and dimension. eg. if someone is wearing a white shirt against a white background, or even if they have white buttons, you shouldn't carve a hole through the person.

            1 vote
            Sign in
            Check!
            (thinking…)
            Reset
            or sign in with
            • facebook
            • google
              Password icon
              I agree to the terms of service
              Signed in as (Sign out)

              We’ll send you updates on this idea

              0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
            • Offer subject pose dection

              While head tilt angle is useful for face detection, it would be incredible to see you support standard 25 (body, no fingers) and 65 (body w/fingers) bone placement for full-body images.

              1 vote
              Sign in
              Check!
              (thinking…)
              Reset
              or sign in with
              • facebook
              • google
                Password icon
                I agree to the terms of service
                Signed in as (Sign out)

                We’ll send you updates on this idea

                0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
              • recognize a barcode or shipping label tracking code for any shipping vendor, and provide a link to the most likely vendor

                Every shipping company and manufacturer chooses a different barcode that supports their needs, but they are not interchangeable, which is by design for automation, but we need to read all the barcodes and categorize them because we cant train our users to figure out which one to scan.

                2 votes
                Sign in
                Check!
                (thinking…)
                Reset
                or sign in with
                • facebook
                • google
                  Password icon
                  I agree to the terms of service
                  Signed in as (Sign out)

                  We’ll send you updates on this idea

                  0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                • OCR Failures

                  1. I'm getting no regions back from about half of the of images I submit for recognition in the attached example file.

                  2. On full document scans, I often see the returned json resembling this:
                  dwoo ,4.Ed co c Ex—co cc äö30a-oF— Z U) r-x.l (DC—a- 0 00 o o O O

                  Can anyone suggest possible causes / solutions to either of these issues?

                  1 vote
                  Sign in
                  Check!
                  (thinking…)
                  Reset
                  or sign in with
                  • facebook
                  • google
                    Password icon
                    I agree to the terms of service
                    Signed in as (Sign out)

                    We’ll send you updates on this idea

                    0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                  • Identify report format

                    There is a need to use Computer Vision/OCR with report images. For that to work well it would be great if the API could make a guess/determination on the layout of the report. Today OCR returns boundingBox information which is helpful but if the API would identify field names from data values as well as header vs. detail fields that would be amazing!

                    1 vote
                    Sign in
                    Check!
                    (thinking…)
                    Reset
                    or sign in with
                    • facebook
                    • google
                      Password icon
                      I agree to the terms of service
                      Signed in as (Sign out)

                      We’ll send you updates on this idea

                      0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                    • Regional availability - Please add support for Taiwan region

                      Please add support for Taiwan region. Thanks

                      1 vote
                      Sign in
                      Check!
                      (thinking…)
                      Reset
                      or sign in with
                      • facebook
                      • google
                        Password icon
                        I agree to the terms of service
                        Signed in as (Sign out)

                        We’ll send you updates on this idea

                        0 comments  ·  Language Support  ·  Flag idea as inappropriate…  ·  Admin →
                      • Font Size & Weight

                        For each word found, note the size of the font as a percentage relative to the largest font found within the image. Or some other similar means. This would allow users to find headings, sub-headings, etc...

                        The font weight (boldness and contrast) of each word would also help to identify headings and important text relative less important text within the image.

                        2 votes
                        Sign in
                        Check!
                        (thinking…)
                        Reset
                        or sign in with
                        • facebook
                        • google
                          Password icon
                          I agree to the terms of service
                          Signed in as (Sign out)

                          We’ll send you updates on this idea

                          0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                        • Vision fails to return tag for specific image

                          I encountered a strange issue using vision when trying to retrieve tags. This post works as expected:
                          POST https://westcentralus.api.cognitive.microsoft.com/vision/v1.0/analyze?visualFeatures=Categories,Tags&details=Celebrities&language=en HTTP/1.1
                          Content-Type: application/json
                          Host: westcentralus.api.cognitive.microsoft.com
                          Ocp-Apim-Subscription-Key: ••••••••••••••••••••••••••••••••

                          {"url":"https://image.shutterstock.com/z/stock-photo-portrait-of-a-man-holding-a-cute-mixed-breed-dog-isolated-over-white-99972185.jpg"}

                          but this image does not return any tags: https://thumbs.dreamstime.com/z/man-holding-unique-dog-2045365.jpg

                          There is no error returned. Any ideas what is going on?

                          1 vote
                          Sign in
                          Check!
                          (thinking…)
                          Reset
                          or sign in with
                          • facebook
                          • google
                            Password icon
                            I agree to the terms of service
                            Signed in as (Sign out)

                            We’ll send you updates on this idea

                            2 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                          • Object recognition for retail

                            I would like to be able to match objects in shelf. This functionality has great value for retail, on stock availability, face count, etc

                            5 votes
                            Sign in
                            Check!
                            (thinking…)
                            Reset
                            or sign in with
                            • facebook
                            • google
                              Password icon
                              I agree to the terms of service
                              Signed in as (Sign out)

                              We’ll send you updates on this idea

                              0 comments  ·  Custom/Sample Images  ·  Flag idea as inappropriate…  ·  Admin →
                            • School Students Handwriting not able to Recoginze .

                              Hi, I am from Bangalore. I tried few students answer sheet using Computer Vision API. Most of the Answer sheet handwriting not able to recognise by API. Can you improve the algorithm or let us know how to train the data. Thank you.

                              1 vote
                              Sign in
                              Check!
                              (thinking…)
                              Reset
                              or sign in with
                              • facebook
                              • google
                                Password icon
                                I agree to the terms of service
                                Signed in as (Sign out)

                                We’ll send you updates on this idea

                                0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                              • Support PDF input in OCR function

                                As stated in the title, I'd like to see the OCR function support PDF files. The only supported formats right now are images like JPG, PNG, etc.

                                4 votes
                                Sign in
                                Check!
                                (thinking…)
                                Reset
                                or sign in with
                                • facebook
                                • google
                                  Password icon
                                  I agree to the terms of service
                                  Signed in as (Sign out)

                                  We’ll send you updates on this idea

                                  0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                                • Add support to Berber languages

                                  The Berber language is a very old language.

                                  There are many benefits of supporting this languages.

                                  For example, there are a lot of very old books that one may want to digitize and index in order to allow millions of Berber-speakers to access to those resources online for example.

                                  There is a lot of effort to do that, but the first step is the have a tool that can perform Optical Character Recognition.

                                  1 vote
                                  Sign in
                                  Check!
                                  (thinking…)
                                  Reset
                                  or sign in with
                                  • facebook
                                  • google
                                    Password icon
                                    I agree to the terms of service
                                    Signed in as (Sign out)

                                    We’ll send you updates on this idea

                                    0 comments  ·  Language Support  ·  Flag idea as inappropriate…  ·  Admin →
                                  • Recognize columns of data/tabular data

                                    As shown in my attached image, there are two columns, but it recognizes only one of them. Strangely, the heading of column 2 is paired with data from column 1.

                                    4 votes
                                    Sign in
                                    Check!
                                    (thinking…)
                                    Reset
                                    or sign in with
                                    • facebook
                                    • google
                                      Password icon
                                      I agree to the terms of service
                                      Signed in as (Sign out)

                                      We’ll send you updates on this idea

                                      0 comments  ·  Custom/Sample Images  ·  Flag idea as inappropriate…  ·  Admin →
                                    • Increase width/height limits on OCR operation to handle Letter/A4

                                      The Image dimension limitation is a problem.

                                      A4 is 210 x 297 mm. At 300 dpi that’s 2480 x 3508 Pixels and 8.7 megapixels. In grayscale as a JPEG of text the page is around 1-1.5MB.

                                      Letter is Image 8.5 x 11 in. At 300 dpi that’s 2550 x 3300 Pixels and 8.42 megapixels. As a grayscale JPEG the size is slightly smaller than A4 (obviously).

                                      The image dimension limit is a bit…silly? Unless there is some overarching technical reason you should increase the maximum dimensions to 3510 x 3510 – so scanned 300 DPI grayscale A4 pages can be…

                                      2 votes
                                      Sign in
                                      Check!
                                      (thinking…)
                                      Reset
                                      or sign in with
                                      • facebook
                                      • google
                                        Password icon
                                        I agree to the terms of service
                                        Signed in as (Sign out)

                                        We’ll send you updates on this idea

                                        0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                                      • Handwriting almost never recognized

                                        I tried it with several handwritten pieces, none of them worked.

                                        1 vote
                                        Sign in
                                        Check!
                                        (thinking…)
                                        Reset
                                        or sign in with
                                        • facebook
                                        • google
                                          Password icon
                                          I agree to the terms of service
                                          Signed in as (Sign out)

                                          We’ll send you updates on this idea

                                          0 comments  ·  Custom/Sample Images  ·  Flag idea as inappropriate…  ·  Admin →

                                          We have just implemented this capability into Computer Vision. Please continue to share your experience and feedback with us.

                                          Check out:
                                          • The new handwriting OCR demo: https://www.microsoft.com/cognitive-services/en-us/computer-vision-api (scroll down to handwriting)
                                          • Documentation: https://docs.microsoft.com/en-us/azure/cognitive-services/computer-vision/home#RecognizeText
                                          • SDKs: Python, Windows, Android
                                          API reference:
                                          o https://westus.dev.cognitive.microsoft.com/docs/services/56f91f2d778daf23d8ec6739/operations/587f2c6a154055056008f200
                                          o https://westus.dev.cognitive.microsoft.com/docs/services/56f91f2d778daf23d8ec6739/operations/587f2cf1154055056008f201

                                        • 2 votes
                                          Sign in
                                          Check!
                                          (thinking…)
                                          Reset
                                          or sign in with
                                          • facebook
                                          • google
                                            Password icon
                                            I agree to the terms of service
                                            Signed in as (Sign out)

                                            We’ll send you updates on this idea

                                            0 comments  ·  API  ·  Flag idea as inappropriate…  ·  Admin →
                                          ← Previous 1 3 4
                                          • Don't see your idea?

                                          Feedback and Knowledge Base