Manage Images
  • 07 Mar 2023
  • 8 Minutes to read
  • Dark
    Light
  • PDF

Manage Images

  • Dark
    Light
  • PDF

Supported Files Types

  • BMP
  • JPEG
  • PNG (recommended)
  • Pascal VOC: Available only for Object Detection projects. Pascal VOC is a file type for images labeled in LandingLens or an external application. This file format includes the label details and tells the platform where a label is on the associated image.

Image Upload Limitations

Number of Images Required for a Project

You must upload and label at least 10 images before training a Model.
 
Since every Project is different, there is no recommended number of images for the project. The optimal number of images for your Project generally depends on these factors:
  • How elaborate the object to identify is
  • How many types of objects there are
  • How the object looks

For example, a phone with a cracked screen would require fewer images than a printed circuit board with several intricate parts.

Images Are Resized During Model Training

When you train your Model, LandingLens resizes all images to make training faster and more effective. This resize does not affect the original images. This means that the original high-resolution images are always available in the Data Browser.

  • When you use Fast Training (the default Train option), all images are auto-resized according to your Project Type.
  • When you use Advanced Train(Train with Custom Options), all images resize to the width and height you provide, with the following restrictions:
    • The sizes of the width and height cannot exceed 1500px.
    • Each Project Type has a maximum area that cannot be exceeded. See Maximum Image Size for Model Training for more information on image size areas.
Note:
If you have large images that include small objects to detect, these objects can be lost due to the automatic resize. You can use Advanced Train to choose the resize width and height manually. Or, you can resize images before you upload them to your LandingLens Project.

Maximum Image Size for Model Training

Note:
Anomaly Detection is only available to legacy "classic" workflow users.

Each Project Type has a maximum image size for Model training:

  • Classification: 1500x1500px
  • Object Detection: 1500x1500px
  • Segmentation: 1024x1024px
  • Anomaly Detection: 512x512px

The maximum image size is measured by the area, not width and height. For example, the maximum image size for Object Detection is 1500x1500px. To get the area, multiply 1500 times 1500, so the equation looks like this: 1500 x 1500 = 2,250,000. If you want to resize an Object Detection image, the area cannot exceed 2,250,000px.

  • A resize of 1000x600px will work because the area of this resize is 600,000px, which is fewer than 640,000px.
  • A resize of 1500x1501px will not work because the area of this resize is 2,251,500px, which is greater than 2,250,000px.

The table below shows the maximum area (in pixels) for each Project Type:

Project TypeMaximum Image Size (in pixels)Maximum Area (in pixels)
Classification1500x15002,250,000
Object Detection1500x15002,250,000
Segmentation1024x10241,048,576
Anomaly Detection512x512262,144

Overview: Upload Options

LandingLens offers the following methods for uploading images to your Project. Click the hyperlink for each method to learn more.

  • Drag and drop: Select the files or folders you want to upload, and drag and drop them into an Upload pop-up window or directly into the Project.
    Drag and Drop Images into the Project
  • Upload images from a directory: Some Upload pop-up windows offer the option to select files from a directory. Select the files or folders you want to upload and click Open.
    Select Images to Upload
  • Load sample data: If you haven't added any images to a Project, you have the option to upload a sample dataset. To add the sample images, click Load Sample Data and select a dataset. The available datasets are based on the Project Type.
    Upload Sample Data
  • Take photos with your webcam: Click the Webcam option and take photos with your computer webcam.
    Upload Photos from Your Webcam
  • Upload pre-labeled images: Upload images in the Pascal VOC file format.

Upload Images to a New Project vs. a Project that Already Has Images

If you haven't uploaded images to a Project yet, all of the upload options appear directly on the Project page.

Upload Options for Projects that Don't Have Images YetAfter you've uploaded images to a Project, click the Upload icon to see the upload options. Or, you can drag and drop images directly into the Project page.

If a Project Already Has Images, Click the Upload Icon to Upload More

Drag and Drop Images into a Project

To drag and drop images directly into a Project:

  1. Open the Project you want to upload images to.
  2. If you haven't uploaded any images to the Project yet, drag and drop files directly into the Project. LandingLens immediately uploads the images.
    Drag and Drop Images into the Project
  3. If you've already uploaded images, click the Upload icon.
    If a Project Already Has Images, Click the Upload Icon to Upload More
  4. Drag and drop files into the Upload pop-up window.
    Drag and Drop Images
  5. Click Upload.
    Upload Images
  6. LandingLens uploads the images to your Project.

Upload Images from a Directory

To upload images from a directory:

  1. Open the Project you want to upload images to. 
  2. Open the Upload pop-up window:
    • If you haven't uploaded any images to the Project yet, click the Drop to Upload icon.
      Open the Upload Pop-Up Window
    • If you've already uploaded images, click the Upload icon.
      Click the Upload Icon
  3. Click the upload box to search for and select the images you want to upload.
    Click to Select Images
  4. A preview of your images displays. To confirm your upload, click the Upload button.
    Upload Images
  5. LandingLens uploads the images to your Project.

Upload Sample Images to a New Project

You can upload a preconfigured dataset if you haven't added any images to a Project. The available datasets are based on the Project Type. To give you more options to experiment, some datasets have labeled images, and some don't.

To add the sample images:

  1. Open the Project you want to upload images to.
  2. Click Load Sample Data.
    Load Sample Data
  3. Select a dataset.
    Select a Dataset
  4. LandingLens uploads the images to your Project.

Upload Images from Your Webcam

To take photos with your webcam and upload them to LandingLens:

  1. Open the Project you want to upload images to. 
  2. Open the Upload pop-up window:
    • If you haven't uploaded any images to the Project yet, click Use WebCam.
      Use WebCam
    • If you've already uploaded images, click the Upload icon and select Turn on Webcam.
      Click the Upload Icon
  3. The first time you upload images from a webcam, your might be prompted to grant LandingLens access to your webcam.
    Allow Access to Your Webcam 
  4. Click the Capture button each time you want to take a photo with your webcam.
    Take a Photo
  5. (Optional) If you also want to upload images from your computer, click Turn off Webcam. You can then drag and drop images into the Project, and select images from a directory.
    Turn Off Webcam
  6. A preview of your images displays. To confirm your upload, click the Upload button.
    Upload Images

Upload Pre-Labeled Images

You can upload images that have been labeled in an external application. These pre-labeled images must be in Pascal VOC file format. Pascal VOC (Visual Object Classes) is a file format that includes the label details of its paired image. It essentially tells the platform where a label is on the associated image. 

Before you upload pre-labeled images, you must unzip the file; LandingLens will not accept zipped files.

Note:
The Pascal VOC file type is only available for Object Detection Projects.

To upload pre-labeled images:

  1. Open the Project you want to upload images to. 
  2. Open the Upload pop-up window:
    • If you haven't uploaded any images to the Project yet, scroll down to the bottom of the page and click Click Here.
      Open the Upload Pop-Up Window
    • If you've already uploaded images, click the Upload icon.
      Click the Upload Icon
  3. You can drag and drop images into the Project, and select images from a directory.
  4. A preview of your images displays. If an image has a corresponding XML file, LandingLens captions it as Pascal Voc.
  5. To confirm your upload, click the Upload button.
    Preview and Upload Pre-Labeled Images
  6. The pre-labeled images display in the Data Browser.
    Labeled Images

Upload Images After Model Training

LandingLens will automatically generate Predictions on images uploaded after Model Training. You will be able to see these Predictions in the Data Browser. This feature will save you time from retraining your Model after you upload new images to your dataset.

In the unlikely event that the Model does not predict an image successfully, the status Unpredicted will display on the image.

"Unpredicted" Image Status 

What if an image does not display any Predictions and has the status "Unpredicted"? If this occurs, the Model did not find anything to detect.

Notes:
  • If you have not already trained your Model, or your Model did not successfully complete its training, the Model will not make any Predictions.
  • It is recommended to label uploaded images, even if the Model's Predictions are accurate.

Upload Folders

LandingLens simplifies folder structures and flattens all files. Let's say you have one folder (Folder A) and two subfolders (Folder B and C). If you upload Folder A, LandingLens will upload all images from Folders A, B, and C, as if they came from the same folder.

LandingLens Flattens Files 

However, if you upload Classified Media to a Classification Project, LandingLens will ignore any subfolders. Referring to the example above, if you upload Folder A, only the images in Folder A will be uploaded; the images in Folders B and C will not be uploaded.

LandingLens Ignores Subfolders for Classified Media Uploads

Delete Images from a Project

Let's say you accidentally uploaded some images to your Project that you do not want to keep. You can remove those images from your Project. To do this:

  1. Select the images you want to remove. 
    • To select a specific image, hover over it and click the checkbox in the upper-left corner. After you've selected one image, you can click anywhere on another image to add it to the selection.
      Click Individual Images
    • To select a range of images, hover over the first image and click the checkbox in the upper-left corner. Press Shift and select the last image in the range you want to select.
      Press the "Shift" Key to Select a Range of Images
    • To select all images or a page of images, click Select and choose either Select All or Select Page.
      Select All Images or Select All Images on the Page
  2. Click Options in the action bar near the bottom of the screen and select Remove from Dataset. (You may need to scroll to see this option.) The selected images are deleted from your Project.
    Remove the Selected Images from the Project
 

Was this article helpful?

What's Next