Blue LED illumination generally provides the highest edge contrast for small features on reflective metals, while white LED illumination supports general dimensional inspection and natural color imaging [S2][S4].
The practical selection depends on material, surface finish, camera type, and defect geometry: blue for precision metalwork, red for reflective or shiny parts, and white for plastics, printed circuit boards, and color-based inspection [S3][S4][S8].
Choose Illumination by Surface and Measurement Feature
Machine-vision lighting should maximize contrast between the measured edge and its surrounding background while remaining stable across part size, orientation, placement, and ambient-light changes [S2].
White light is the general-purpose choice because it produces balanced illumination across a wide range of materials and suits plastic parts, printed circuit boards, and matte mechanical components [S4]. It is also the appropriate option when a color camera must reproduce natural colors [S3].
Blue and white illumination are common choices, and the appropriate lighting depends on the material, surface finish, and measurement requirements. Illumination choice affects image contrast, edge-detection accuracy, and measurement repeatability. Because lighting choice affects measurement repeatability, illumination should be selected according to the material, surface finish, and measurement requirements.
Red light is ideal for inspecting reflective or shiny materials. Lighting geometry still governs what the camera sees: backlighting emphasizes outlines, dark-field lighting reveals scratches and engravings, and ring lighting gives broadly uniform illumination [S3].
Compare the Three LED Colors Against Decision Criteria
Blue and white are common illumination types in optical measurement systems, red light is ideal for inspecting reflective or shiny materials, and the appropriate choice depends on the application.
These statements are not contradictory: the useful measure is the contrast generated at the particular defect and background, not a universal ranking by wavelength.
Match Lighting Geometry to the Edge Being Measured

Backlighting places the object between the camera and source, making only the outline, recesses, holes, and fill level visible; this arrangement is appropriate when dimensional outline rather than surface texture is the measurement target [S3].
Dark-field illumination directs light onto the object at a shallow angle so that only areas at a specific angle to the camera reflect light, leaving most of the image dark while revealing edges, scratches, engravings, and other surface defects. Different lighting techniques can be used to address various inspection tasks.
For shiny components, compare red with blue under the actual camera and lens because reflections, coatings, and surface curvature can reverse the expected contrast ranking [S4][S8].
Use White Light for General and Color Inspection
White LED light provides uniform illumination for plastics, printed circuit boards, electronic components, and matte mechanical parts, and it is the practical default for general dimensional measurement [S4].
White illumination can be used with a color camera to reproduce natural colors. A color camera under white illumination can reproduce natural colors.
A 2025 study reported diminished contrast sensitivity after exposure to white LEDs at 450 lx, a 45-degree incidence angle, and an 80 cm source distance, with measurements made using Chart-2020 software. That finding concerns human exposure and should not be used as a machine-vision edge-contrast coefficient, but it supports controlling optical exposure rather than treating illuminance as unlimited.
For machine imaging, verify saturation before reducing wavelength or increasing intensity. Record the target edge contrast, camera exposure, illuminance, and repeatability under the installed geometry, then compare red, white, and blue using the same camera settings and representative production parts.
Apply Red Light to Reflective and Shiny Materials

Red illumination is identified as an effective option for inspecting reflective or shiny materials, where it can separate surface response more effectively than a broadly specified white source [S8].
Red light should not be selected merely because the workpiece appears red or because the camera is monochrome. The test condition is whether the selected wavelength produces a repeatable difference between the edge and the adjacent background after the coating, roughness, curvature, and specular reflections are included.
A low-vision discussion notes that daytime-visible wavelengths extend from 360 nm at the near-ultraviolet boundary to 830 nm at the near-infrared boundary, and that red and blue features can differ in perceived edge contrast because the eye does not resolve all spectral channels equally [S6]. That perceptual relationship is not a substitute for sensor measurements, but it explains why visible sharpness and camera-recorded edge contrast may not match.
Red illumination is consequently best considered a surface-specific tool: test it against blue on polished metal, tooling, and coated components, and retain the color that gives the larger grayscale separation and lower repeatability variation.
Control Optical Variables Before Changing LED Color
The dominant lighting variables are geometry, spatial pattern, wavelength, and filtering, so changing LED color before controlling these four factors can conceal the real cause of poor edge contrast [S2].
Camera spectral response must also be included. The same LED wavelength can produce different grayscale levels across sensors, and a light source selected for human appearance may not deliver the strongest signal in the camera’s useful band [S2]. A monochrome camera with strong blue or red sensitivity can use either specialized source, while a color camera under white light remains the practical choice for natural-color inspection [S3].
For dimensional metrology, keep the geometry and camera fixed while swapping colors. Capture at least 20 repeated images from each representative production state, then compare edge localization uncertainty, grayscale gradient across the boundary, and measurement repeatability. If contrast changes after the part rotates, the illumination is not robust even if it looks excellent in one sample position.
Pixel calibration remains a separate step from illumination choice. A useful reference for pixel-to-mm calibration in machine vision measurement explains why improving edge definition must still be followed by spatial calibration before dimensional results are accepted.
Final Selection Rules for Stable Edge Measurement

Select blue LEDs for fine edges on polished, machined, moulded, or micro-mechanical metal; select red LEDs for reflective or shiny surfaces that need a differentiated reflection; select white LEDs for general measurement, mixed materials, and natural-color inspection [S3][S4][S8].
Use backlight for outlines, dark field for scratches and shallow surface defects, and ring or horizontal-ring illumination for homogeneous or edge-reflection tasks [S3]. Keep the wavelength decision tied to measured contrast and repeatability rather than a general claim that one color always produces sharper edges.
The next verification record should contain five entries for each candidate color: illuminance, incidence angle, camera exposure, grayscale edge contrast, and repeatability across representative parts. That record will expose whether wavelength solved the inspection problem or merely masked a geometry, filter, or camera-response issue.