Buffer Manage

DWT collects all the input files and stores them in a buffer. This section of the guide will explore the different aspects of the buffer, including its features and abilities as well as its limits. First, let’s explore the limits of the buffer and how to use the local cache:

Memory Limits

When DWT works in Service mode, it can run in 32bit and 64bit and the data is stored raw in DIB format. Before version 15.0, it’s 32-bit by default and that means it can utilize no more than 2 GB of physical memory. In version 15.0 onwards, 64-bit has been made the default option on a 64-bit OS and that means there is no limitation on how much memory it can use.

On the other hand, if DWT works in WASM mode, it can utilize no more than 2 GB of memory as WebAssembly only supports 32-bit pointers at present. However, the data is stored in JPG format or (for Black & White images) PNG format which are much smaller than DIB .

In order to control the size of the data, DWT can limit the number of images allowed in buffer with the property MaxImagesInBuffer .

Disk Caching

This section applies to desktop browsers running in service mode.

The data DWT deals with are images which take up lots of space as they are stored raw in DIB format. For example, one A4 paper scanned in 300 DPI takes around 24MB in memory which means 2GB of physical memory can only store no more than 85 of these images. As more images are processed, more memory gets used which may pose a threat to other programs on the machine. Due to this, the disk cache function was added. After enabling disk caching, most images will be temporarily cached on the disk, while keeping some active images in the memory to maintain high performance.

The disk caching feature is enabled by default and can be disabled by setting IfAllowLocalCache to false .

You can also set how much memory DWT can use before images start to be cached. By default, 800MB is used. You can change it using the property BufferMemoryLimit .

NOTE

All cached data is encrypted and can only be accessed by DWT .

When DWT is unloaded (like when the browser tab refreshes or closes), the cached data is destroyed and removed from the disk automatically.

Although you can scan and load as many images as you like, you may want to handle them in smaller volumes when doing further processing. For example, when dealing with extremely large volumes, you should not try to upload or save the images all at once as that would be too time consuming and prone to errors.

Rearranging the images within the buffer

DWT offers a number of methods to help you shift the images within the buffer.

NOTE

You can also just drag the images around in the viewer to change their positions.

• Use the MoveImage() method:

/* Assuming there are 5 images in the buffer, the following code changes the 1st image to be the 3rd. The sequence changes to 12034 from 01234*/
DWObject.MoveImage(0, 2);

• If you are looking to simply switch two images in the buffer, use SwitchImage() :

// Switch the positions of the 1st and the 3rd images
DWObject.SwitchImage(0, 2);

• Lastly, you can use one of the following methods to remove one or multiple images from the buffer:

// Removes a single image specified by index
DWObject.RemoveImage(1);
//Removes all images in the buffer
DWObject.RemoveAllImages();
// Removes the first 3 images from the buffer
DWObject.SelectImages([0, 1, 2]);
DWObject.RemoveAllSelectedImages();

Tagging Images

DWT offers the ability to place tag(s) on image(s) which helps putting them into groups. A tag is a string that describes the corresponding grouping. The following demonstrates how to tag images:

/* Tags the first two scanned images with the classification "title" to indicate that they belong to the title section of the book being scanned. The 3 images after the first two belong to chapter 1 of the book so they were tagged correspondingly */
DWObject.TagImages([0, 1], "title");
DWObject.TagImages([2, 3, 4], "chapter1")

If you are scanning in a new batch of documents and would like to put the same tag on all of them, you can do so using the SetDefaultTag() method.

/* The next batch of documents are the medical records of patient123 so instead of tagging each document after they are scanned, tag them all as such. */
DWObject.SetDefaultTag("patient123");
DWObject.AcquireImage();

To filter the images in the buffer by a tag, use FilterImagesByTag() .

To clear the tags on a specific image, use ClearImageTags() .

Image Data

The buffer also provides access to a wide array of info for each image. Here is the full list of the info available via the buffer:

• Bit depth, which helps to identify the pixel type of the image: GetImageBitDepth()
• Height of the image in pixels: GetImageHeight()
• Width of the image in pixels: GetImageWidth()
• Horizontal resolution of the specified image: GetImageXResolution()
• Vertical resolution of the specified image: GetImageYResolution()

The following is only supported in desktop browsers and in Service mode.

• Size in bytes of the specified image according to input dimensions: GetImageSize()
• Size, but this time you specify an image format and DWT calculates the size in bytes depending on that image format: GetImageSizeWithSpecifiedType()
• Size in bytes of selected images according to input dimensions: GetSelectedImagesSize()

Status Reporting

Now that we’ve covered the general image info like size and dimensions, let’s address some of the less intuitive but still useful pieces of information that our buffer can offer:

• To get the total number of images in the buffer, use HowManyImagesInBuffer

• The index of the image currently active in the buffer is represented via the CurrentImageIndexInBuffer property, which you can also use to set the current image

• In addition to each image having a unique index, each image also has a unique image ID. To find the image ID of an image at a certain index, use IndexToImageID() and use ImageIDToIndex() if vice-versa

NOTE

Unlike the index which changes when the position of the image changes, the image ID stays unchanged until the image is no longer in the DWT buffer.

• If a scanned image comes out a bit skewed, and you are not able to put that specific page through the scanner again, one potential solution would be to rotate the image by how much it is skewed, which can be easily done using GetSkewAngle() .

Image selection is a powerful tool that can be utilized if you want to carry out an action but only on specific images in the buffer. We will go over some of the methods used to set the selected images in the buffer:

• SelectImages() is the main method used here. All you need to do is feed it the array of image indices you want to select
• SelectAllImages() simply selects all the images in the buffer
• SelectedImagesIndices returns the indices of the selected images as an array

The following is only supported in desktop browsers and in Service mode.

• The method GetSkewAngleEx() is used to get the skew angle of a specified rectangle on the specified image
• Each image in the buffer also has an internal URL which you can use to reference it, accessible via GetImageURL() and GetImagePartURL() . This could come in handy for certain situations, like retrieving and showing the images in your own viewer.

Note

GetImageURL() returns a URL that has the normal scheme http:// or https:// while GetImagePartURL() returns an internal URL that has the scheme dwt:// . The latter can only be accessed by other DWT APIs.

Blankness Detection

Only supported in desktop browsers and in Service mode.

The buffer also comes with the ability to detect if a scanned image is blank or not. This is easily done using one of the two following methods: IsBlankImage() or IsBlankImageExpress() . The 2nd method is faster but not as accurate as the first.

If you believe an image should be blank but IsBlankImage() or IsBlankImageExpress() is returning false , you can read BlankImageCurrentStdDev (a read-only property) for that image and then set a bigger value to BlankImageMaxStdDev , which sets the “limit” at which the buffer considers the image blank. Then when you try again, that image will be regarded as blank.

Buffer Event Callbacks

The buffer also provides a few event callbacks that take place under specific conditions. You can use these callbacks to trigger a certain action once this event takes place. Here is an overview of the events and what each represents:

• OnBitmapChanged This event is triggered when the data in the buffer changes. Specifically when

  • New image(s) enters the buffer
  • Old image(s) gets removed
  • Existing image(s) gets edited

• OnTopImageInTheViewChanged This event is triggered when the first visible image in the viewer changes.

• OnIndexChangeDragDropDone This event is triggered when you drag and drop images in the viewer to change their positions.