What Is a Fundus Camera? Types, Optics and Embedded Integration
A fundus camera is a low-power microscope with its own light source and a camera attached, built to photograph the retina through the pupil of a living eye.
That one sentence hides most of the engineering. The instrument has to put light in and take an image out through the same small opening, hold focus across a wide range of patient prescriptions, and do it fast enough that the patient does not blink.
This post covers what is inside the instrument, the fundus camera types you will meet, and what changes when you build one rather than buy one.
What Is a Fundus Camera?
It is an ophthalmic imaging device that produces a photograph of the retina, and it differs from an ordinary camera in having its own illumination path arranged so the light never blinds the image.
A retinal camera and a digital fundus camera are the same thing under different names, the second emphasising that the image lands on a sensor rather than on film. An eye fundus camera is the same device again in consumer-facing copy.
The comparison people reach for is fundus camera vs ophthalmoscope. An ophthalmoscope lets one clinician look at one retina, live, and nothing is recorded. A fundus camera captures an image that can be stored, compared against last year’s, sent to a grader in another city, or fed to software. In a fundus camera vs ophthalmoscope decision, the question is whether you need a record or a look.
That recording ability is why the instrument sits at the centre of screening. A retinal screening camera in a clinic photographs eyes faster than a specialist can examine them, and the images are graded later, which is the whole economics of a retinal screening camera programme.
What Is Inside a Fundus Camera
The fundus camera parts divide into three groups: the optics that form the image, the illumination that makes it possible, and the electronics that capture it. Understanding the parts is the quickest route into fundus camera optics, because each group exists to solve one problem the eye creates.
The objective sits closest to the eye and forms the retinal image. Everything downstream depends on it.
The illumination source provides two kinds of light in most instruments: a continuous low-level beam for alignment, often near infrared so the pupil does not react to it, and a flash for the capture itself. Fundus camera illumination is therefore two systems sharing one path.
The beam separator is the part that makes the optics unusual. Illumination and imaging share the same pupil, so they are separated inside the instrument, classically by a holed mirror that reflects the illumination outward while letting the image pass through its centre. That holed mirror arrangement produces annular illumination: the light enters as a ring rather than a disc, and annular illumination is what keeps corneal reflections out of the picture.
The focusing group shifts the image plane to compensate for the patient’s refraction. Its range is the instrument’s dioptre compensation, and a typical range covers around fifteen dioptres either side of neutral.
The sensor and electronics sit at the end of the chain and turn the formed image into data.
The patient interface is the chin rest, forehead bar, and joystick or positioning stage, and it does more for fundus camera image quality than most specification sheets admit. A head that moves costs you the shot.
How a Fundus Camera Captures the Retina
Align: The operator brings the instrument to the eye and centres it on the pupil, watching a live preview lit by the low-level beam. Set the illumination ring wrong against the pupil edge and reflections wash out the frame.
Focus: The focusing group is adjusted until the retinal detail sharpens, which compensates for the patient’s prescription rather than for distance.
Capture: The flash fires and the sensor records one frame. In most instruments, the working distance is fixed by the objective, so fundus camera working distance is a design parameter rather than something the operator varies.
Review: The image is checked, and if the eye blinked or the alignment drifted, it is taken again.
So how a fundus camera works comes down to that alignment and focus discipline, and the instrument’s job is to make both easy enough that an ordinary technician succeeds. This is what imageability rate measures: the proportion of eyes from which a gradable image is actually obtained. A camera with excellent optics and a poor imageability rate is worse in a screening clinic than a modest camera that succeeds every time.
Mydriatic and Non-Mydriatic Fundus Cameras
A mydriatic fundus camera works on a dilated pupil. A wider opening means more light in and out, better peripheral coverage, and a more forgiving alignment. The cost is clinical: drops take twenty minutes to work, blur the patient’s vision for hours, and mean they cannot drive home.
A non-mydriatic fundus camera works through a natural pupil, usually in a dimmed room where the pupil has widened on its own. Nothing is instilled, the patient leaves immediately, and throughput rises sharply. The cost is optical: a smaller aperture, less light, and less tolerance for alignment error.
Screening is dominated by the non-mydriatic type for exactly that reason. A fundus camera for a diabetic retinopathy programme photographing hundreds of people a day cannot dilate them all.
The compromise most instruments make is near-infrared illumination. The pupil does not constrict to light it cannot see, so the operator can align at leisure and the eye stays open.
Tabletop, Handheld and Portable Fundus Cameras
A tabletop fundus camera is the clinic standard. A heavy base, a chin rest, and a positioning stage give stability, and stability is what produces consistent images. It is also the format with room for the largest optics.
A handheld fundus camera goes to the patient. Bedside, paediatric, domiciliary, and field screening all need something you can carry, and there is no chin rest to hold the head still, so the instrument has to tolerate movement and an untrained operator.
A portable fundus camera sits between the two: light enough to move between rooms or sites, often on a small stand, with more optical capability than a handheld and less bulk than a tabletop.
The trade-off is consistent. Stability and optical size fall as portability rises, and the imageability rate in unskilled hands falls with them. That is why handheld instruments lean so heavily on autofocus and on generous alignment aids, and why a fundus camera for glaucoma follow-up, where images are compared year on year, is usually a tabletop.
Standard and Wide-Field Fundus Cameras
Thirty and forty-five degrees are the long-standing standards, with forty-five the usual screening choice because two or three frames cover the clinically important area. Sixty degrees widens it further.
A wide field fundus camera goes beyond that, into the range where a single frame covers most of the posterior pole and some of the periphery. The attraction is obvious: fewer frames, less repositioning, shorter appointments.
What widening costs is less obvious and worth stating. The same sensor spread across more degrees resolves fewer pixels per degree, so small lesions land on fewer pixels. Illumination uniformity also gets harder, because the edges of a wide field are further from the axis than the centre. A wide field fundus camera buys coverage and sacrifices resolution and evenness.
Neither is better. A screening programme hunting small lesions wants pixels per degree; a programme hunting peripheral pathology wants coverage.
What to Look for When Specifying a Fundus Camera
Field of view: because it sets everything downstream, including how many frames per eye and how much detail each one holds.
Dioptre compensation range: because a range that does not reach your patient population means some eyes cannot be imaged at all.
Imageability rate: in the hands of the staff you actually have, not in the hands of an ophthalmic photographer at a trade show.
Fundus camera working distance: which determines how close the instrument sits to the face and how comfortable the examination feels.
Connectivity: A DICOM fundus camera drops into a hospital imaging workflow without a translation layer, and a DICOM fundus camera is usually a requirement rather than a preference once the buyer is a hospital rather than a clinic.
Fundus camera image quality is assessed on your own patients. Manufacturer sample images are taken on cooperative eyes in good conditions.
Building One: The Embedded Camera Inside
An OEM fundus camera programme starts from a board-level camera module rather than a boxed camera. You design the optics, the illumination geometry and the patient interface, and the module supplies the imaging front end: sensor, lens mount, image processing and a standard interface over USB or MIPI CSI-2.
What you gain is fit. The form factor is yours, the field of view is yours, the software is yours, and the instrument can be exactly as wide, as light or as specialised as your market needs. What you take on is everything the boxed instrument hid: the optical design, the illumination path, the alignment aids, and the eye-safety and regulatory work that goes with them.
An embedded fundus camera design also changes the supply question. A module you can buy for years, in the same configuration, with the same drivers, is worth more to a medical device programme than a slightly better sensor that disappears at the next refresh.
Vadzo supplies that imaging front end as board-level modules in USB and MIPI CSI-2 form, with customization for optics, connector, and firmware. The sensor, focus, and interface specifications, and which module suits which field of view, are covered in our guide on fundus photography and camera module selection.
To be clear about the boundary: the module is the imaging front end. The retinal optics, illumination path, pupil alignment, and eye-safety engineering remain the device maker’s responsibility.
Fundus Camera Modules by Imaging Mode
A finished fundus camera usually supports more than one imaging mode, and the mode decides the module more than the resolution does.
Imaging mode | Module | Why this one |
Colour fundus, routine exams and screening | Falcon-1335CRA (USB), Bolt-1335CRA (MIPI) | 13MP colour with VCM autofocus. The default for a tabletop instrument. |
Handheld and portable capture | Falcon-1335CRO (USB), Bolt-1335CRO (MIPI) | Same sensor with optical stabilisation, which matters when no chin rest is holding the head still. |
Fluorescein angiography time series | Falcon-258CRA (USB) | Phase-detect autofocus refocuses in a single step per frame, which a sequence capture needs. |
High-contrast FA and difficult lighting | Falcon-821CRH (USB), Falcon-830CRH (USB), Bolt-830CRH (MIPI) | HDR holds bright leakage and dark ischaemic zones in one frame. |
ICGA, SLO and NIR reflectance | No colour filter array, so every pixel collects the full near-infrared signal. |
Two points about that table. The monochrome row is not an afterthought: fundus camera parts lists differ between a colour instrument and one that also does indocyanine green work, because the NIR modes need a sensor with no Bayer mask in front of it. And the OIS rows earn their place only in handheld designs, since a chin-rest instrument has already solved the stability problem mechanically.
The 20MP AR2020 modules sit outside this table deliberately. They are high-resolution general medical imaging parts rather than modules positioned for retinal work, and they belong in a fundus camera discussion as an option to evaluate rather than as a recommendation.
Frequently Asked Questions
What is a fundus camera?
A fundus camera is a low-power microscope with its own illumination and an attached sensor, built to photograph the retina through the pupil. Answering what a fundus camera is properly means naming the thing that makes it unusual: light goes in, and the image comes out through the same small opening, so the two paths are separated inside the instrument. A retinal camera or digital fundus camera is the same device, and an eye fundus camera is the same thing in plain language. The fundus camera vs. ophthalmoscope distinction is exactly that: a record rather than a live look, which is why a retinal screening camera can photograph eyes faster than a specialist can examine them. Vadzo supplies the imaging front end for OEM instruments of this kind rather than finished cameras.
How does a fundus camera work?
The operator aligns on the pupil under a dim alignment beam, adjusts the focusing group for the patient’s prescription, fires the flash, and reviews the frame. Asking how a fundus camera works is really asking about that alignment: the annular illumination enters through the outer ring of the pupil while the image leaves through the centre, classically split by a holed mirror, which keeps reflections out of the picture. Fundus camera illumination usually runs continuously in near-infrared for alignment and flashes in visible light for capture, and those are the fundus camera parts that most affect fundus camera optics design. The instrument’s real job is making that sequence easy enough that an ordinary technician succeeds. That success rate, the imageability rate, matters more in a clinic than any optical specification.
What is the difference between a mydriatic and a non-mydriatic fundus camera?
A mydriatic fundus camera images a pupil dilated with drops, which gives more light, wider coverage, and easier alignment, at the cost of twenty minutes of waiting and hours of blurred vision for the patient. A non-mydriatic fundus camera works through the natural pupil in a dimmed room, so the patient leaves immediately and throughput rises, but the smaller aperture is less forgiving. Screening programmes overwhelmingly use the non-mydriatic type, which is why a fundus camera for diabetic retinopathy service can photograph hundreds of people in a day. Instruments bridge the gap with near-infrared alignment; since the pupil does not constrict to light, it cannot see. A fundus camera for glaucoma follow-up more often justifies dilation, because year-on-year comparison rewards the better image.
Which type of fundus camera should I choose?
Start with where the patient is, because that narrows the fundus camera types faster than any specification. A tabletop fundus camera gives the stability and optical size a clinic wants; a handheld fundus camera goes to the bedside or into the field; and a portable fundus camera sits between them for multi-site work. Then decide coverage: forty-five degrees is the screening standard, while a wide-field fundus camera covers more retina per frame and gives up pixels per degree and illumination evenness to do it. Check the dioptre compensation range against your patient population and confirm the fundus camera working distance suits your examination room. For an ophthalmic imaging device going into a hospital, connectivity usually decides it, and a DICOM fundus camera saves a custom integration. Judge fundus camera image quality on your own patients rather than on manufacturer sample images.
Can we build our own fundus camera instead of buying one?
Yes, and an OEM fundus camera programme is common when no boxed instrument fits the market you are selling into. You start from a board-level camera module supplying the imaging front end over USB or MIPI CSI-2, then design the optics, illumination, and patient interface around it. What you gain is control of form factor, field of view, and software; what you take on is the optical design and the eye-safety and regulatory work. An embedded fundus camera design also rewards long-term module availability, because a medical programme cannot absorb a sensor change mid-life. Vadzo supplies the modules and the customization for optics, connectors, and firmware, while the retinal optics, illumination, and safety engineering stay with you.
