Define What Must Fit Inside the Image

Begin with the inspection drawing rather than a preferred focal length. Record the whole area that must remain visible, including locating features, orientation changes and position uncertainty. Specify horizontal and vertical dimensions at a named object plane. For moving products, include the positions at which the image may actually be captured.

Field of view describes the object area seen by the imaging system. Working distance describes the separation between an agreed lens reference and that object plane. Edmund Optics’ imaging fundamentals also separates these from sensor size, pixel size and magnification. Keep those inputs distinct in the RFQ.

All example dimensions in this guide are hypothetical. They illustrate the calculation sequence and are not a guaranteed configuration or SUPERIOR product specification.

Check the Camera Aspect Ratio

Required coverage must fit in both directions. A 4:3 active sensor used at one magnification produces a corresponding rectangular field for ordinary, low-distortion imaging. If the object and sensor aspect ratios differ, one dimension will normally include additional scene area, or the output will need cropping.

For the example, 180:135 and 12.8:9.6 are both 4:3. A horizontal calculation therefore produces the required vertical coverage as well. If a later camera mode crops the image to a different ratio, repeat the calculation using its actual active area and retained pixels.

Do not assume that resizing an output image gives the same result as capturing more optical detail. State whether the algorithm uses raw pixels, binned pixels, a cropped region or a scaled image.

Calculate Magnification and Object-Space Sampling

Use physical sensor dimensions for magnification and pixel count for object-space sampling:

Magnification m = sensor width / horizontal FOV

Object-space sampling = horizontal FOV / horizontal pixel count

The same relationships apply vertically when the dimensions are paired consistently. For the illustrative camera:

Whether 4.55 pixels is sufficient depends on the task, contrast, blur, motion, lighting and algorithm. Establish the required pixel coverage through the application’s validation method. Avoid declaring a universal pixels-per-defect rule or treating a subpixel algorithm claim as a complete system accuracy specification.

Edmund Optics’ resolution note connects sensor sampling and magnification. The calculations above are original worked examples using those relationships, with their assumptions stated explicitly.

Estimate Focal Length, Then Check the Actual Lens

A commonly used first-pass approximation for conventional imaging is:

f ≈ sensor width × working distance / horizontal FOV

Using 12.8 mm, 400 mm and 180 mm gives an estimate of about 28.4 mm. This is a starting point for comparing available lenses. It is not a promise that a nominal 28 mm lens will produce the exact field at a mechanically measured 400 mm.

The approximation depends on simplified geometry. Entrance-pupil and principal-plane locations, focusing behavior and distortion can matter. In its focal-length and field-of-view guide, Edmund Optics explicitly cautions against treating these estimates as final mechanical design values.

For shortlisted models, obtain field-of-view data or a lens-specific calculation at the intended distance. Verify the supplier’s working-distance reference, minimum focus distance, focus travel and tolerances. Keep adjustment allowance in the installation instead of freezing the machine drawing around the approximate value.

Check Practical Constraints Before Choosing a Model

The lens specifications guide provides the next step for reading MTF, aperture and distortion evidence under comparable conditions.

Decide Whether Conventional or Telecentric Optics Fit the Task

Use conventional fixed-focal-length optics when their coverage, image quality and perspective behavior meet the inspection requirement. Consider telecentric optics when apparent size changes with object distance are important to dimensional measurement. Confirm the actual magnification, working distance, sensor coverage and specified telecentric performance of the proposed model. Edmund Optics describes the underlying perspective distinction in The Advantages of Telecentricity.

These are engineering screening routes, not a guarantee that a lens family solves the application. Review the machine vision lens portfolio and telecentric lens options only after defining the task and camera.

Run a Coverage-to-Acceptance Trial

Use representative objects at the permitted positions and heights. Keep camera processing documented; sharpening, resizing and automatic exposure can conceal what changed between trials. Capture center and corner features, then repeat the near and far object conditions without silently refocusing.

Record the actual field dimensions and measured installation distance, the focus-lock state, aperture, light settings, exposure, gain and processing. Identify the characteristic that determines acceptance: readable characters, defect detection, edge position or dimensional repeatability. Compare against the application limit rather than the visual impression of a sharp photograph.

For procurement, separate three outcomes: the lens fits; the image covers the required area; and the complete system passes the task. Only the last outcome supports the final application approval.

Request a Lens Review With the Geometry Attached

Send the camera or sensor datasheet, required horizontal and vertical coverage, working-distance envelope, object-height range and smallest relevant feature. Include a machine sketch showing obstructions and lighting access. Discuss your machine vision lens selection with SUPERIOR using those inputs so candidate models can be reviewed against a defined installation.