Remove Periodic Interference with Notches
The microscope adds regular bands at a few repeatable frequencies.
Read the full lesson and open its transcript below. A narrated video will be added after production.
What you will learn
The microscope adds regular bands at a few repeatable frequencies.
A practical question
The microscope adds narrow, evenly spaced bands to every scan. In the Fourier view, that periodic interference produces localized peaks. A notch filter can weaken those selected components while leaving many other frequencies alone. This is more selective than blurring every small detail.
The useful idea
For a real-valued image, the relevant frequency peaks appear as symmetric partners in the transform. We should inspect both, and choose narrow regions around them. Then reconstruct the image and compare it with the untouched specimen, especially where the specimen itself has a repeating texture.
Work through the example
Widen the notch experimentally. The artificial bands may fade further, but a genuine pattern with similar spacing can fade too. That is why we check known structures in several scans and record the exact filter mask. Don't rely on the cleaned image alone to explain what was removed. A regular scanner stripe can produce concentrated peaks that are easier to target than a broad range of natural detail. Mark the pair of suspected peaks, then remove only a narrow neighborhood and reconstruct the scan. Compare a known good area and a real fine feature with their original versions. If the feature changes, your notch may overlap information you wanted to keep. The peaks identify repeating structure, but their presence does not prove the scanner caused it. That judgment needs another capture or evidence about the specimen. Now check the idea on a concrete example: Identify symmetric peaks from periodic interference and attenuate a narrow neighborhood around them. A possible trap is worth naming: An overbroad notch can erase genuine repeating structure. The answer follows from that evidence: A notch filter suppresses selected frequency regions rather than all fine detail.
Limits and next step
The result may also show ripples around abrupt transitions if the filtering mask changes sharply. Next, we'll learn to recognize those ripples before anyone mistakes them for features of the specimen.
Worked example
Identify symmetric peaks from periodic interference and attenuate a narrow neighborhood around them.
Quick check: test your understanding
Try it yourself: Consider this case: Identify symmetric peaks from periodic interference and attenuate a narrow neighborhood around them. Explain how you would check the result. In your answer, address this limitation: An overbroad notch can erase genuine repeating structure.
Check the reasoning
A notch filter suppresses selected frequency regions rather than all fine detail. An overbroad notch can erase genuine repeating structure.
What is the main idea behind remove periodic interference with notches?
A notch filter suppresses selected frequency regions rather than all fine detail.
How is it used in the example?
Identify symmetric peaks from periodic interference and attenuate a narrow neighborhood around them.
What limitation should be checked?
An overbroad notch can erase genuine repeating structure.
Further reading
Digital Image Processing, 4th edition — Rafael C. Gonzalez and Richard E. Woods.
Original ToolsInVision teaching examples. Editorial approach · Suggest a correction
Accessible referenceLesson 43 transcript308 words · Narration ready for the upcoming videoOpen transcript
The microscope adds narrow, evenly spaced bands to every scan. In the Fourier view, that periodic interference produces localized peaks. A notch filter can weaken those selected components while leaving many other frequencies alone. This is more selective than blurring every small detail.
For a real-valued image, the relevant frequency peaks appear as symmetric partners in the transform. We should inspect both, and choose narrow regions around them. Then reconstruct the image and compare it with the untouched specimen, especially where the specimen itself has a repeating texture.
Widen the notch experimentally. The artificial bands may fade further, but a genuine pattern with similar spacing can fade too. That is why we check known structures in several scans and record the exact filter mask. Don't rely on the cleaned image alone to explain what was removed. A regular scanner stripe can produce concentrated peaks that are easier to target than a broad range of natural detail. Mark the pair of suspected peaks, then remove only a narrow neighborhood and reconstruct the scan. Compare a known good area and a real fine feature with their original versions. If the feature changes, your notch may overlap information you wanted to keep. The peaks identify repeating structure, but their presence does not prove the scanner caused it. That judgment needs another capture or evidence about the specimen. Now check the idea on a concrete example: Identify symmetric peaks from periodic interference and attenuate a narrow neighborhood around them. A possible trap is worth naming: An overbroad notch can erase genuine repeating structure. The answer follows from that evidence: A notch filter suppresses selected frequency regions rather than all fine detail.
The result may also show ripples around abrupt transitions if the filtering mask changes sharply. Next, we'll learn to recognize those ripples before anyone mistakes them for features of the specimen.