Yes. Baseline comparison software takes new photos or a video walkthrough of a space, compares them against an earlier record of the same space, and reports what changed, what got damaged, and what went missing. It is sold as AI inspection, AI damage detection, or visual change detection. RapidEye is built around exactly this: it keeps a baseline for every room it has seen, reads each new set of photos or walkthrough video against it, and returns a short list of findings instead of a folder of images. The thing to check in any tool is whether it actually compares against the space's own history, because most photo AI only judges one picture at a time.
The question usually comes from someone who already has the pictures. Cleaners photograph every turnover, attendants photograph finished rooms, a landlord has a move-in video, a venue filmed the hall before the event. The record exists. What does not exist is a person with the time to put this week’s photos next to last week’s and spot the difference. That comparison is what this category of software does.
The three things it reports
Every finding a baseline comparison tool produces falls into one of three kinds. A tool that only does one of them is a narrower product than the question implies.
Damaged
Something is newly broken, stained, or marked
A surface that was clean in the baseline now shows a defect: a stain, scratch, crack, burn, hole, or tear.
New stain on sofa cushion, living room Crack in shower glass, bath 2Missing
Something in the baseline is no longer there
An object the space is supposed to contain does not appear in the new capture.
Lamp missing from left nightstand 5 of 6 dining chairs presentChanged
Something is present but not as it should be
Moved furniture, a setup that drifted from standard, or a cleaning miss the baseline did not have.
Hot tub cover left unlatched Bed made with wrong throw, primary“Missing” is the one that separates real baseline comparison from everything else. You cannot see that a lamp is gone by looking at one photo of a nightstand. You can only see it by knowing the nightstand had a lamp last time.
How it works, without the buzzwords
01
BaselineAn earlier, known-good record of each room: the guest-ready state, the move-in condition, the pre-event hall.02
New captureThe photos or walkthrough video you already take, from the same rooms, on the next visit.03
ComparisonThe software matches rooms and objects across the two records and looks for differences.04
FindingsA short list of damaged, missing, and changed items, each with the before and after image.Two things decide whether the output is useful. First, the baseline has to be good: a clean, complete record of the space taken when it was in the condition you want to hold it to. Second, the new capture has to cover the same things. Good tools tolerate different camera angles, lighting, and who took the pictures; no tool can compare a surface that never appeared in the new capture. That is why a fixed walkthrough route matters more than a fixed tripod position.
Video helps for the same reason. A 60 to 90 second continuous walkthrough covers every surface the camera passes, so the comparison is not limited to the angles someone remembered to photograph. RapidEye reads both photos and walkthrough video against the same baseline; our guide to what AI can check in a video walkthrough lists what it finds room by room.
Baseline comparison vs. ordinary photo AI
Many inspection and task apps now advertise AI on photos. Most of it checks a single submission at the moment it is taken. That is useful, but it is not what the question asks for.
| Question you need answered | Single-photo AI | Baseline comparison |
|---|---|---|
| Is this photo blurry, dark, or the wrong room? | Yes | Yes |
| Does this room look clean and made up? | Yes | Yes |
| Is that scratch new, or was it there last month? | No | Yes |
| Is anything missing that should be here? | No | Yes |
| Which visit did this damage first appear on? | No | Yes |
| Has the setup drifted from standard over time? | No | Yes |
The “which visit” row is where most of the money is. Knowing damage exists is not enough to charge for it; you have to show when it appeared. A tool that compares every visit against the last one establishes that on the day it happens. Our piece on damage attribution with back-to-back bookings walks through why.
Who uses it, by type of space
The method is the same everywhere: a known-good record, a later capture, a list of differences. What changes is who hands the space to whom.
| Space | Baseline | Comparison capture | What the findings decide |
|---|---|---|---|
| Vacation rentals | Guest-ready record | Cleaner photos or video after each checkout | Damage claims, cleaner quality, missing inventory |
| Hotels | Brand-standard room setup | Attendant photos of the finished room | Room release, re-clean, maintenance tickets (AI room inspection) |
| Long-term rentals | Move-in inspection | Move-out or mid-lease walkthrough | Deposit deductions vs. normal wear and tear |
| Event venues | Pre-event condition | Post-event walkthrough | What the organiser is charged (pre and post event inspections) |
| Homes rented for shoots | Pre-shoot walkthrough | Post-wrap walkthrough | What the production pays for (renting to film productions) |
Construction progress tools also compare captures over time, but they answer a different question (how far along is the build, does it match the plan) and are not designed to report a missing lamp or a new stain.
What it cannot see
Baseline comparison sees what the camera sees, and nothing else. Plan for a human or a separate check on:
- Smell. Smoke, pets, and mildew do not photograph.
- Function. A dishwasher that looks normal and does not run, a TV that will not turn on unless the capture shows it on.
- Hidden surfaces. Inside closed drawers, under rugs, behind furniture that was not moved.
- A missing baseline. If the first good record of a room was never taken, the first comparison has nothing to compare against. The fix is to capture one now.
The honest value is not that a machine catches everything. It is that your team gets a short list of flagged changes, each with its before and after, instead of scrolling every photo or none of them.
Questions to ask any vendor
- Does it compare against the space’s own history, or judge each photo on its own?
- Does it report missing items, or only visible damage?
- Does it read video walkthroughs, or only still photos?
- Can it read the photos you already take in your existing app, or do your staff need a new one? RapidEye reads from the platforms operators already use; see integrations.
- Does every finding show the before and after image and the date it first appeared?
- Can a person confirm or dismiss each finding before it goes to a guest, tenant, or client?
- Does it keep the original files with their capture time, so the record holds up in a dispute?
Quick FAQ
What is this kind of software called?
Baseline comparison, visual change detection, or before-and-after inspection software. In property operations it is usually sold as AI inspection or AI damage detection. The defining feature is that it compares a new capture against an earlier record of the same space, rather than judging a single photo on its own.
Can it work from video instead of photos?
Yes, and video is usually the better input. A short continuous walkthrough covers every surface the camera passes, so the comparison is not limited to the angles someone remembered to photograph. RapidEye reads both photos and walkthrough video against the same baseline.
Do the photos have to be taken from the exact same angle?
Not exactly, but the capture has to cover the same things. Software that matches images by room and object tolerates different angles and lighting; what it cannot recover is a surface that never appeared in the new capture. A fixed walkthrough route matters more than a fixed tripod position.
How is this different from the photo AI built into inspection apps?
Most built-in photo AI checks one submission at capture time: is the photo blurry, is the bed made, does the room look clean. It does not know what the room looked like last week, so it cannot tell you a lamp is gone or a scratch is new. Baseline comparison is history-aware by definition.
Who uses baseline comparison software?
Anyone who hands a space to someone else and needs to know its condition when it comes back: vacation rental and hotel operators between guests, landlords at move-in and move-out, event venues before and after an event, owners renting a home for film shoots, and facilities teams checking rooms on a schedule.
Will it catch everything?
No. It sees what the camera sees. Smells, a broken appliance that looks normal, the inside of a closed drawer, and damage under an object fall outside what a photo can show. The value is that your team gets a short list of flagged changes instead of every photo.