Applications of filters in visual robots

Introduction to Optical Filters

MOK Optics, a professional supplier specializing in optical filtering, provides full-spectrum filter solutions from deep ultraviolet to far-infrared. This article focuses on MOK Optics’ product line, exploring the applications of bandpass filters, narrow bandpass filters, IR-cut filters, and UV-cut filters in robot vision systems.

I. Applications of Optical Filters in Robot Vision

1.1 The Qualitative Change in Lighting Environment

In laboratory environments, lighting is often uniform, controllable, and predictable. However, when robots are deployed in real-world applications, they face drastically different scenarios:

Industrial Inspection Scenarios: In underground utility tunnels, oil and gas fields, or mines, robots need to operate in environments with extremely low illumination (as low as 0.1 lux), high dust levels, or strong thermal radiation (such as infrared stray light from high-temperature equipment).

Smart Home and Security Scenarios: Smart door locks or robot vacuums need to withstand strong midday sunlight, yellow streetlights at night, and even malicious human interference (such as screen photos or supplemental lighting).

Faced with these complex environments, ordinary vision sensors are affected by the environment, resulting in low efficiency. In such cases, our filters are needed to optimize the robot’s signal reception and signal quality.

II. The Technological Foundation of MOK Optics Filters

2.1 Core Parameters: Defining the “Hard Power” of Filters

To understand why filters can change the success or failure of machine vision, we first need to understand several key technical indicators that MOK Optics strictly controls during the manufacturing process:

Center Wavelength & Bandwidth (CWL & FWHM): The center wavelength defines the midpoint of the spectrum that the filter allows to pass through. For machine vision, bandwidth (FWHM, or half-width at half-maximum) is particularly critical. MOK Optics offers a range of options, from ultra-narrowband (FWHM ≤ 10nm) for precision laser detection to broadband (FWHM > 50nm) for general illumination. Narrowband filters precisely lock onto signals and eliminate stray light when robots identify structured light or lidar signals at specific wavelengths.

Optical Density (OD): This is the core metric for a filter’s “blocking power.” MOK Optics filters achieve cutoff depths up to OD6, meaning only one in a million (10⁻⁶) of stray light can penetrate. This extremely high blocking power is crucial for obtaining clean signals in Raman spectroscopy analysis or robot vision under intense laser conditions.

Durability (Environmental Adaptability): Robots often operate in environments with vibration, high humidity, high temperature, or dust. MOK Optics employs advanced hard coating technology and precise substrate material engineering to ensure that filters do not peel or crack under extreme conditions, maintaining stable optical performance.

III. Practical Applications of MOK Optics’ Main Filters in Robotic Scenarios

3.1 IR-cut and UV-cut Filters

IR-cut filters and UV-cut filters are the most basic and important combination in robot vision.

Principle and Function: The visible light range for humans is approximately 380nm-780nm. However, CMOS/CCD sensors are exceptionally sensitive to near-infrared (NIR) and ultraviolet light. Without intervention, during daytime shooting, leaves reflect strong near-infrared light, causing strange color casts in the image (leaves appear white, color distortion). UV-cut filters effectively block ultraviolet light with wavelengths below 380nm, avoiding excessive blue noise; while IR-cut filters block infrared light with wavelengths above 780nm, ensuring that the spectrum received by the sensor is strictly limited to the visible light range, thus guaranteeing the authenticity of color reproduction.

Robotics and Intelligent Security:

Daytime Mode: In intelligent inspection robots or security cameras, the IR-cut filter works to block infrared light, allowing the robot to see images in true colors, facilitating color label identification or equipment status assessment.

Nighttime Mode: When light levels decrease, the IR-cut filter automatically moves away (or switches to another piece of fully transparent glass), allowing infrared illumination to reach the sensor, achieving black-and-white night vision. This “dual-mode switching” is achieved using a precise IR-cut filter. MOK Optics’ IR-cut filters have high transmittance in the visible light band and extremely high infrared cutoff depth, ensuring clear and usable images in both day and night modes.

3.2 Bandpass and Narrow Bandpass Filters: Anti-interference “See-Through Eyes”

In industrial machine vision and specific inspection scenarios, bandpass filters and narrow bandpass filters are crucial for eliminating ambient stray light and extracting effective information.

“Penetration Techniques” in Industrial Inspection: As mentioned earlier, visible light struggles to penetrate dusty oil and gas fields or mines. MOK Optics’ far-infrared bandpass filters are used in the gimbal cameras of robotic dogs. These filters allow only specific far-infrared wavelengths of light to pass through (wavelengths that can penetrate dust and sensitively capture temperature differences), thus clearly displaying heat sources in the environment (such as temperature anomalies at equipment cracks) in the image.

3D Vision and Structured Light: Many robots rely on structured light or Time-of-Flight (ToF) for depth recognition. Under sunlight, these wavelengths suffer from strong background light interference. MOK Optics’ narrowband filters precisely match the laser emission wavelength, allowing the emitted spectrum to return efficiently to the sensor while significantly attenuating sunlight interference from other wavelengths (above OD4). This enables robots to perform stable face recognition or obstacle avoidance operations even in strong outdoor light.

3.3 Long-pass and Short-pass Filters: The “Sorters” of Spectral Screening

Longpass filters allow light with wavelengths longer than the cutoff wavelength to pass through. UV-cut filters are essentially a type of long-pass filter. These filters are crucial in machine vision sorting applications.

IV. Conclusion

MOK Optics, with its comprehensive product line—from basic IR-cut and UV-cut filters to high-precision narrowband and long-pass filters, and deep customization capabilities for specific wavelengths—provides the robotics industry with eyes that can “see clearly and accurately” in various complex lighting environments.