Select faults in Sensoft Vision

Applies to

Sensoft Vision 2 and later

In graphs

You can select faults by dragging. Selected faults are shown in blue in the graphs, hovering over one shows how many are selected out of the total (see Figure 1). To add faults to the current selection hold the Ctrl key while dragging, to subtract faults hold the Shift key. Note that you can select faults also on page Statistics, e.g. in the histogram graphs.

With selected faults you can do many things: browse through them (with the Left and Right key, or by scrolling the mouse wheel on the photo), see how they correlate (on page Statistics), or, if you think they are fake faults and have the permission, exclude or even delete them.

With the Select faults combobox

Figure 1: Page Results with the Select faults combobox highlighted. The selected faults are marked in blue in the upper graph and the count in visible in the label.

On page Faults there is a combobox, highlighted in Figure 1, that allows you to select photos by type. A type is often a category of the AI model, but can be also all faults with a particular feature, e.g. those with photo.

Some of the most useful can be selected from the drop-down list, but there are many more, and they can be combined, e.g. Selected and Uncertain selects from the selected faults those that have a feature in the class Uncertain. You can put "No ", "not " or similar words in front of a category, e.g. no photo or Without photo selects all faults that have no photo. You can combine two or more categories with comma (",") or the words "and" and "or". E.g.: Faults 100 to 200, not selected, and real. If your string is valid, it gets pretty-printed and stays black. If it is invalid, it is marked red.

Fault - Select the fault with that fault number. Type "Fault 33" or simply "33" to go to fault nr. 33. Type "Faults 33 to 55" or "33-55" to select those faults.

Critical - The faults that affect whether the spool is a FAIL. Works only for spools with alarm criteria, which can be set when using Advanced criteria. Works only for spools measured with Sensoft Vision 2 or later.

Small - If a spool has hundreds of faults, the file size can get an issue. Each fault with photo uses approximately 0.5 MB on disk. This category selects the smallest faults, so that when deleted, the spool occupies no more than 50 MB on disk. Use "Small and with photo" to select only the small faults with photo.

Time - Select faults based on the when they happened. E.g.: "From 15:20 to 15:30", "At 3:20 PM" and "15:20" all work and do the same.

Photo - Select faults that have a photo.

Excluded - Select the excluded faults, i.e. those that have been measured, but assessed not to be real (either by the user, or if specified in the Criteria by the AI model). The opposite is Included.

Selected - The faults that are already selected. Useful only when combined with other types, e.g. Selected and Uncertain finds within the selected faults those with a feature that the AI model could not recognize.

All - Select all faults. The contrary is None, which deselects all.

Real - The AI model has found on the photo at least a feature in the category Lump, Neck-down, Damage, Irregular enamel or Scratch.

Unsure - The AI model has found no features in the Real categories, but at least a feature in category Uncertain.

Fake - The AI model has found either no feature or only features in the categories Hair, Dust or Mosquito. Thus Fake means with photo, but neither in category Real nor Unsure.

Results - Select the faults that the AI model could analyze successfully. Normally this means all faults with photo. Mostly used in the form "No results".

Nothing - Select faults with photo, where the AI model has found no feature. Feauters are displayed with bounding boxes on the photo.

Lump - AI category: The AI model has found on the photo a feature in the category Lump, i.e. a diameter increase like e.g. a blister.

Neck-down - AI category: Diameter decrease.

Damage - AI category: Damaged wire.

Irregular enamel - AI category: Uneven enamel or uneven copper.

Hair - AI category: A hair or hair-like dust particle.

Dust - AI category: A dust particle.

Mosquito - AI category: An insect.

Scratch - AI category: A scratch on the filament.

Uncertain - AI category: Features the AI model detected, but was not able to categorize.