VAPAR

Welcome to our Knowledge Portal
< All Topics
Print

Reviewing the AI Outputs​

Watch the following video which demonstrates how you can review the data that has been generated by the platform AI:

The video doesn’t answer your question? Try reading the text following the video to find out more information.

Results will only appear here, on the Survey page, once the uploaded CCTV file has completed all stages of processing. If you cannot see your results on the Survey page, check the processing status of your files on the Files page or the Processing Videos list on the Upload Files page. The processing status definitions can be found in the Processing files section in the ‘Uploading with Full Platform Access’.

Each row of the Survey table represents the results of a single uploaded CCTV video. The table is automatically populated based on the information shown in the uploaded video. There are also warning indicators that may be shown on the Survey page as shown in red below.

The meaning of these indicators are given below.

Individual file actions

The Actions provided to the far right of the Survey table will also assist in viewing your results

The descriptions of each of the actions shown in the Actions columns are given below.

The frame-by-frame view

To view and review the analyzed data for each video, select the ‘View’ button on the Files or Survey page, or ‘Show Details’ on the map.

Editing your videos

Watch the following video which demonstrates how you can edit the metadata:

The video doesn’t answer your question? Try reading the text following the video to find out more information.

Video metadata

VAPAR uses optical character recognition (OCR) to populate the inspection metadata fields. A range of different camera software is used to hard-code text onto the inspection videos, please get in touch with VAPAR for information on the options available to optimise this for VAPAR.

To edit the metadata for your video, click the edit icon to launch the editing window. 

Using this editing window you are able to edit the following details:

●      Date of inspection

●      Location

●      Asset ID

●      Start node

●      End node

●      Chainage

●      Diameter

●      Camera direction

●      Material

●      Water level warning

●      Loss of vision warning

When beginning your review, it is good practice to start with a quick check to see if anything requires adjustment or manual entry. In some instances not all the data may be picked up from the inspection title frame. If parts of your metadata are consistently not carrying across automatically, please get in touch with us at VAPAR so we can investigate the possible causes and solutions.

The example above is missing the following fields: Street, End Node, Direction. The quickest way to check if this information is available is viewing the Title Frame page by toggling All Frames and Include Hidden Frames to on.

 

In this example, we can edit the metadata by pressing the pencil icon and adding in: Cassino Dr, 61063, Upstream. The updated metadata box shown below.

 

Click the Save Changes button to ensure your changes are saved. Once saved, the updated details will be displayed in the Analyse table, and will also be reflected in all exported results.

Defects

 

To view larger versions of the individual analysed frames thumbnails, simply click on the thumbnail. This will open the raw analysed frame in the main frame view.

If your organisation is on a basic subscription the defects will not be automatically classified, and remain blank in the defect view until the classification is manually assigned or the frame is hidden.

The structural and service scores will not display until all frames have been classified or hidden. The results will show similar to below:

Filtering analysed defect frames

Watch the following video which demonstrates how you can change how you view the frames:

The video doesn’t answer your question? Try reading the text following the video to find out more information.

You are also able to view the results by toggling between the options shown in the top right side of the Analyse table.

The following are explanations of the viewing options:

      1. The Show Defects will show only those images that have been associated with a unique defect or pipe feature from the uploaded video. By default, this option is selected.
      2. The Show All will show all the analysed frames of the uploaded video that have been classified as important or unique.
      3. The Show Hidden and Unhidden frames will show all the hidden analysed frames of the uploaded video. Hidden frames are frames that have been tagged as such either by a user or by our algorithms. Common reasons for tagged frames as hidden is if they are duplicate defect frames of an already flagged defect, or the image is irrelevant to the condition assessment of the pipe.

To view all extracted frames in a video select both “Show All” and “Show Hidden and Unhidden frames” options.

Editing analysed defect frames

Watch the following video which demonstrates how you can edit defects:

The video doesn’t answer your question? Try reading the text following the video to find out more information.

When you select a frame, a yellow border will appear, the main frame view will show the selected frame, and the frame edit box will be filled with the values associated with this frame.

The two main reviewing actions that will be typically be taken are:

      1. Changing a defect classification
      2. Hiding or unhiding an image based on its relevance.

Both of these actions can be done from the frame selection box.

Change the defect classification by clicking on the dropdown and selecting a different defect.

Hide or unhide an image by clicking on the eye icon.

You can edit the additional frame values and details using the edit box.

The descriptions of each of the values in the frame editing box are given below.

Once changes are made within this box, press Save Changes to save the changes made.

Next Survey Workflow
Table of Contents
Shopping Basket