For counting cartons at the dock, barcode scanning is usually the cheapest place to start, because shipper cartons can carry a printed barcode at no extra tag cost. RFID fits best when tags are already applied at source and whole pallets move fast through a door, while cameras at a fixed pass point fit conveyor loading where you want an image of every carton as evidence.
Each method fails in a different way. This guide compares them on cost per carton, infrastructure, what breaks reads, speed, evidence and fit for Indian 3PLs and FMCG distributors.
How each method works
Barcode
A scanner reads a printed code. On shipper cartons this can be an ITF-14 symbol, which GS1 describes as a barcode for “corrugated boxes” that can be printed “on a label or directly on a corrugate” (GS1 India). Scanning can be handheld at the door or fixed over a conveyor. Each code needs line of sight.
RFID
A passive UHF tag on each carton or pallet answers a reader through antennas at a dock door or portal. No line of sight is needed. Dock door portals “consist of RFID readers and antennas” (RFID4Ustore).
Camera-based counting
Cameras at a fixed point, usually over a conveyor, detect each carton, read its printed code or barcode, and keep an image. Image-based barcode readers have long offered image archiving so users can “see what the reader sees” (Logistics Management).
Side-by-side comparison
| Factor | Barcode | RFID (passive UHF) | Camera-based |
|---|---|---|---|
| Cost per carton | Low if printed on the carton at packing; label and printing cost if added later | A tag on every carton; see sourced prices below | No per-carton consumable; uses existing printed codes |
| Infrastructure | Handheld scanners, or fixed readers over a conveyor | Readers, antennas, often shielding at each door | Camera arch or mount, lighting and a computer at each pass point |
| Line of sight | Needed for every code | Not needed | Needed for the carton face being read |
| What hurts reads | Torn, smudged or wrinkled labels; glare from shrink wrap; label in the wrong place | Metal and liquids; tag orientation; dense pallets; stray reads from nearby doors | Overlapping or stacked cartons; hidden faces; poor lighting; damaged print |
| Single-file conveyor | Good with fixed readers | Good | Good |
| Wrapped pallet through a door | Scan the pallet label; cases inside are not seen | Can read cases inside without breaking the pallet, if products are RF-friendly | Weak; inner cartons are hidden |
| Evidence for disputes | Scan log | Read log | Image and timestamp per carton |
| Setup effort | Low to medium | High: tagging programme, site survey, tuning | Medium: fixed pass point, mounting, setup per site |
Cost per carton
This is where the three methods differ most.
RFID. Every carton needs a tag, and it is consumed when the carton is sold. In a 2022 report, RFID Journal said tags “are often sold at retail prices of 10 or 20 cents per tag”, while one low-cost tag averaged around 4 US cents when bought in Asia in quantities of 500,000 or more. Preprogrammed versions cost about 11 cents each in quantities of 100,000 (RFID Journal, 2022). A label with a printable surface roughly doubles the cost of a bare tag (RFID Journal). Prices change, so get current quotes.
Barcode. GS1 UK notes that ITF-14 works well printed directly on cardboard boxes and recommends printing it on two sides (GS1 UK). If the brand already prints it at packing, the dock pays nothing per carton. If you must add labels, you pay for labels, printers and labour.
Camera. There is no tag or label to add if cartons already carry a readable code. The cost sits in the hardware at each pass point and in setting it up for your carton range.
What breaks each method
RFID: metal, liquids and stray reads
“Liquids absorb RFID energy”, and “metal reflects RF waves”, which can detune tag antennas and create null zones (Atlas RFID). This matters for beverages, edible oils, personal care and canned goods. In an older RFID Journal answer, the editor said generic tags on RF-friendly goods often read on 100 percent of cases, but on pallets of bottled water the read rate “could be only about 60 percent” because inner cases are hard to reach (RFID Journal, 2009). That answer dates from 2009, so test with your own products.
Dock doors sit side by side, so readers can pick up tags at the next door. Shielding is used to “reduce stray reads from adjacent lanes or doors”, and tag orientation, pallet density and reader placement all affect read rates (RFID4U). Knowing whether a tag is going out or coming in also takes extra design, such as photo eyes or additional antennas (University of Ottawa survey of UHF RFID portals).
Barcode: label condition and scanning discipline
Damaged labels “will not decode properly”. Smudges, ribbon wrinkles, other labels in the quiet zone and glare from shrink wrap or clear tape all cause no-reads (Advanced Automation). With handhelds, limited range means workers “may need to move closer, shift pallets, or repeat scans” (Keyence). The bigger risk is human: a carton loaded without a scan is a carton not counted.
Camera: occlusion and the pass point
A camera can only count what it sees. With tightly stacked boxes, object detection “struggles with overlapping boxes”, and “fully hidden boxes are not recoverable from a single image” (Ultralytics community). The same discussion notes that tracking across video frames helps. In practice, cameras work best where cartons pass one by one, as on a loading conveyor, with cameras on more than one side.
Conveyor versus pallet loading
The way your trucks are loaded often decides the method.
- Cartons on a loading conveyor. All three can work. Fixed barcode readers and cameras use the codes already printed. Cameras add an image of each carton.
- Wrapped pallets through a door. Scan the pallet label at the threshold; one guide’s rule is “if the barcode won’t scan, the pallet doesn’t load” (SG Systems). RFID can read cases inside the pallet when products are RF-friendly. Cameras cannot see inner cartons.
- Hand loading without a conveyor. Handheld barcode scanning is the practical choice. RFID Journal suggests reading case tags as cases go on or off a pallet when inner reads are a problem (RFID Journal).
Evidence when a distributor disputes a load
Barcode and RFID give you a data log: which code, when, at which door. That is strong evidence if your process forces every carton through the check. Cameras add a picture of each carton with its timestamp, which a distributor or auditor can look at without trusting your software. Photos of the load and seal still help with any method. “Scans + EPCIS + photos win; anecdotes lose” (SG Systems). For the full dispute process, see our guide on how to stop shortage and excess disputes at dispatch.
When each option is the better choice
Choose barcode if you load by hand or by pallet, your cartons already carry clean barcodes, and you want the lowest cost.
Choose RFID if the brand already tags cartons at source, products are RF-friendly, and pallets must pass through doors without being broken down. It is also easier to justify on reusable crates or pallets, where one tag is read many times.
Choose cameras if cartons pass single file on a conveyor, the load mixes many SKUs with printed codes, you do not want to tag, and you need an image of every carton for disputes.
Fit for Indian 3PLs and FMCG distributors
- Who pays for tags. For a 3PL, tagging other companies’ cartons on arrival adds a tag and labour cost to every carton. RFID works best when the brand tags at source.
- Mixed client labelling. Different clients print different codes. A method that reads what is already printed avoids relabelling.
- Dock layout. Cameras and fixed readers need a fixed pass point. RFID portals need power, network and tuning at each door (RFID4Ustore).
- Product mix. Beverages, oils and canned goods are hard for RFID. Shrink-wrapped and dusty cartons are hard for barcodes. Stacked loads are hard for cameras.
Where DeepVision Warehouse fits
DeepVision Warehouse from Indus Vision takes the camera route for conveyor loading. Its dock tally is an arch over the loading conveyor with cameras overhead and on both sides. It reads each carton’s printed code and barcode and tallies it in real time against the shipment, uploaded as an Excel file of order lines. It shows ordered versus loaded per carton code with short and excess live, holds unknown cartons aside for an admin, and keeps an image and timestamp for every carton. It works for loading and unloading, runs on site, needs no WMS change and exports XLSX. It was built in an FMCG distribution centre. If your loads move as wrapped pallets or your cartons are tagged at source, barcode pallet scanning or RFID may suit you better. See also our FMCG solutions page.
Frequently asked questions
Is RFID more accurate than barcode for dock counting?
Not automatically. RFID reads without line of sight, but metal, liquids, dense pallets and stray reads all affect it. Barcode is reliable when labels are clean and every carton is scanned. Test both on your own products and docks.
How much does a UHF RFID tag cost?
RFID Journal reported in 2022 that tags often sell at 10 to 20 US cents at retail, with one tag averaging around 4 cents in very large volumes in Asia (RFID Journal). Printable labels and encoding cost more. Ask suppliers for current quotes.
Can cameras count cartons inside a wrapped pallet?
No. A camera only sees outer faces. Hidden cartons cannot be recovered from a single image. Cameras work best where cartons pass one at a time.
Can we combine methods?
Yes. You can scan pallet labels at the door, count cartons by camera or scanner on the conveyor, and weigh the loaded vehicle as a cross-check. Pick the method per flow, not per site.