Add References

This function is only available for data in the IDCube format. For other formats, the user needs to first convert the data files and reference files into the IDCube format. 

Features:

  • Removes artifacts related to the illumination and noise from the detector.

  • Provides a pure spectrum of the sample

Steps:

1.       Open a folder file in IDCube format. Switch to the RGB mode and adjust band/channels/wavelengths. Visualize the spectra of the object. The reflection spectra are severely affected by specular reflection from the light source and at some degree by the dark noise from the detector. 

2.       To correct the image, you will need to have control images recorded under similar conditions. The example folder has two correction files. The white_ref image shown below is recorded from a strip of white reflective material made of Spectralon®. The dark_ref image is recorded by the same setting with closed lens of the imaging system.

3.     To apply the correction files select File Add reference in IDCube format and select one of the three options White, Dark, or White and Dark.

  • White – corrects for white reflection only.

  • Dark – corrects for the dark count only.

  • White and Dark - corrects for white reflection and dark count.

  • Remove corrections – deletes all corrections.

When the White and Dark option is selected, a message box Select the White Reference First will appear. After clicking OK, a pop-up dialoge window Select White Reference File will appear. Choose the white_ref correction file and click Open.

After selecting a white ref file, a message box Now Select the Dark Reference  will appear.   You will be prompted to select a Dark Reference file after clicking OK. Choose the dark_ref correction file and click Open.

The completion of the task will be confirmed by a message box.

The IMAGE DISPLAY and the DIRECTORY INFORMATION panel will be automatically updated. To visualize the spectra, click multispectral mode and select the region of interest. The corrected spectrum shown as a mean spectrum +/- standard deviation is drastically different from the non-corrected (see above). One of the features of the corrected spectrum is a relatively high standard deviation at both ends of the spectrum. This is caused by the limited sensitivity of the detector at the wavelengths shorter than 950 nm and longer than 1650 nm.

Note: A new dataset corrected for the white and dark noise can be treated as a typical IDCube dataset.

Additional Information:

Illumination in hyperspectral imaging systems is often not uniform across the image, generating potential artifacts. This problem can be eliminated by calibrating the image with the white object. Currently, the widely used standard reflectance surface is made of Spectralon® white diffuse reflectance target, although other materials especially in remote sensing can also be used. In addition, imaging detectors used in HSI systems generally have dark current. Dark current is dependent on temperature and is proportional to integration time. To convert raw intensity into reflectance, reference and dark images are taken before or after acquiring sample images. The reference image is taken with a standard reflectance surface placed in the scene, and the dark current is measured by keeping the camera shutter closed.

References:

Lu, Guolan, and Baowei Fei. "Medical hyperspectral imaging: a review." Journal of biomedical optics 19.1 (2014): 010901

Previous
Previous

Open Dataset in IDCube Format

Next
Next

Add Multiple Datasets in IDCube Format