2.80:Detect Frames (Activity)
|
This article is about an older version of Grooper. Information may be out of date and UI elements may have changed. | |
| 2025 | 2.80 |
.png)
view_module Detect Frames is an Activity that locates and identifies frame lines on microfiche card images, enabling the isolation of areas within the frames for further data extraction or processing.
This is the second step in Grooper’s microfiche processing, after an Initialize Card activity has been performed. Documents are arranged on a fiche card in a matrix of rows and columns. The gaps between each document forms a frame of blank space that borders the document. The Initialize Card activity detects these frame locations in a fiche card strip. Once Grooper knows where the frame is around a document, it can go to the next step, Clip Frames, which takes the document image out of the fiche card and stores it as a page in Grooper.
The activity also generates a low-resolution preview of strips of documents on the fiche card. These strips are composed of the tile images organized into subfolders during the Initialize Card activity. But functionally, they are just horizontal strips of the fiche card containing rows of documents. These strip preview images are saved on the strip folder and can be used as the display image during a Review activity to verify frame detection and resolve any frames flagged as having issues.
Version Differences
Detect Frames is a brand new activity in Grooper 2.80. In previous versions, documents from microfiche scans were pre-processed using software local to the scanner. This significantly slows down the speed at which cards can be scanned. Grooper's microfiche capabilities allow the scanner to run at full speed, while Grooper pre-processes them independently. Furthermore, while microfiche scanners do have some image cleanup capabilities, they are nowhere near as robust as Grooper's Image Processing activities. The result is end-to-end microfiche document processing with a faster workflow and cleaner images resulting in more accurate OCR data.
Use Cases
Microfiche is often used for archiving documents. Court records are one example of documents that may be stored on microfiche.
How To: Configure the Activity
Before you begin
The microfiche card must first be scanned directly into Grooper or otherwise imported. The images must then go through the Initialize Card activity to organize the tile images into strip folders.
Set the Card Layout
Under the “General” heading, set the Card Layout settings. Here you specify four important things: How many total rows of documents are on the card (Row Count). How many columns (Column Count). How many rows of documents to expect per strip (Rows Per Strip). And, whether the card has a Reverse Alignment. By default, Grooper assumes the first strip only contains a single row. However, if the last strip contains a single row instead of the first, change the Reverse Alignment setting to “True”
Adjust the preview image DPI and other General settings
You can change the DPI of the strip preview image under Processing Resolution, rotate all tiles by a set degree under Tile Rotation, and how a color or greyscale image is converted to 8-bit black and white under Binarization.
Set the Gutter Detection settings
“Gutter Detection” sets the length and width of the lanes of blank space (called “gutters”) between rows and columns of documents. You can alter their Minimum Horizontal and Vertical Length, the Maximum Gap between documents (or the gutter’s thickness) and their Minimum Thickness. Where the gutters intersect forms the frame border.
Adjust the Page Detection settings
Under “Page Detection”, you can set the Minimum Intensity for each page. Pages are expected to be mostly white (100%). A page with a large amount of black (0%) could indicate a frame detection problem. Anything falling below the value set here will be flagged for human review. You can also set the minimum and maximum Page Size Range, the maximum number of empty cells allowed at the end of a fiche card (Maximum Empty Cells), and whether or not missing frames will be inferred from the set of detected frames (Infer Missing Frames).
Property Details
| Property | Default Value | Information |
| General Properties | ||
| Card Layout | 13 rows X 28 columns | Here you specify four important things: How many total rows of documents are on the card (Row Count). How many columns (Column Count). How many rows of documents to expect per strip (Rows Per Strip). And, whether the card has a Reverse Alignment. By default, Grooper assumes the first strip only contains a single row. However, if the last strip contains a single row instead of the first, change the Reverse Alignment setting to “True” |
| Process Resolution | 75 | This property controls both the DPI (dots per inch) at which frame detection is performed and controls the resolution of the strip preview image. The higher the DPI, the more accurately the activity will detect frame edges and the better the image quality of the preview image will be. However, it will cost processing speed. |
| Tile Rotation | 0 | You can automatically rotate tiles by 90, 180 or 270 degrees by adjusting this property. |
| Binarization | Auto | Binarization converts color images to black and white by "thresholding" the image. Thresholding is the process of setting a threshold value on the pixel intensity of the original image. Pixel intensity is a pixel's "lightness" or "brightness". Essentially, once a midpoint between the most intense ("whitest") and least intense ("blackest") pixel on a page is established, lighter pixels are converted to white and darker are converted to black. Or put another way, pixels with an intensity value above the threshold are converted to white, and those below the threshold are converted to black. This midpoint (or "threshold") can be set manually or found automatically by a software application. The Thresholding Method can be set to one of four ways:
Each method has its own set of configurable properties. For more information on binarization and these methods, visit the Binarize article. |
| Gutter Detection Properties | ||
| Minimum Vertical Length | 16in | This is the minimum length of a vertical gutter, the space running vertically between documents on a microfiche card. |
| Minimum Horizontal Length | 64in | This is the minimum length of a horizontal gutter, the space running horizontally between documents on a microfiche card. |
| Maximum Gap | 2in | This is the maximum gap between documents allowed during gutter detection. |
| Minimum Thickness | 0.25in | This specifics the minimum thickness of horizontal and vertical gutters between the rows and columns of images on a microfiche card. |
| Page Detection Properties | ||
| Minimum Intensity | 90% | Any detected page with an average intensity below the value set here will be flagged for human review. "Intensity" here means how white or black an image is overall, with 100% intensity being pure white and 0% being pure black. Pages should be mostly white space with a little black, accounting for text or images on the page. A page with a significant amount of black (in other words, a low intensity value), could indicate the frame was detected improperly or the page has quality issues. |
| Page Size Range | 4in - 18in | The minimum and maximum dimensions of a page are set here. They will correspond to the smallest and largest edges of pages on a microfiche card. |
| Maximum Empty Cells | 0 | Some microfiche cards have empty cells, meaning there's a missing document in the row and column matrix of documents. Set this value to 0 to force a cell seen as "empty" for human review. If you expect to see empty cells, set this value to the number of empty cells you expect. |
| Infer Missing Frames | False | If enabled, the position of missing frames will be inferred from the position of frames that were detected. |