-
The seller wasn't in the header: a second look at Russian receipts
Re-reading Russian slips in three enlarged bands finds the seller OCR missed. Held-out sellers go from 106 to 113 of 146, none lost, totals unchanged.
-
When OCR reads АО as AD: catching misread legal forms on Russian receipts
The parser now takes the seller even when OCR misreads its legal form. Held-out Russian slips go from 104 to 106 of 146, with none lost.
-
Russian receipts in the browser: a Cyrillic OCR model, hash-pinned, read on your machine
The bookkeeper's page now reads Russian receipt photos with a Cyrillic recogniser in the browser: sellers 116 of 146 on val, the CLI's 117.
-
Telling a Russian receipt from an English one by its misread capitals
The bookkeeper's receipt reader now picks Latin or Cyrillic from the photo: 148 of 151 held-out Russian slips, 570 of 572 held-out Latin receipts.
-
Our bookkeeper got slower every month of the year, and how we stopped it
A bank import took 165 ms in January and up to 620 ms by December. Matching transfers in a date window, not all pairs, cut December to 185 ms.
-
A Cyrillic OCR model read no more totals, but it read the shop's name
Swapping PaddleOCR's East Slavic recogniser into RapidOCR left Russian receipt totals unchanged and took the merchant from 0 to 71% right on held-out slips.
-
Teaching tesseract Russian did not help it check a better engine's receipt totals
Tesseract's Russian model lifts its own vote on Russian receipt photos from 68 to 79 of 145, but adds no total behind RapidOCR's 140.
-
A better OCR engine was losing three receipt dates in ten to a missing space
On 313 held-out Malaysian receipt scans, RapidOCR's glued date and time cost the reader 31% of dates. Splitting them lifted dates from 69% to 94% right.
-
When the receipt loses its total, the card terminal's slip still has it
On a fresh shard of Russian receipt photos, 18 of 25 misses had no total line. The bank slip printed under the receipt fixed five of them.
-
The receipt OCR hang was OpenMP, not the photos
tesseract seemed to hang on Russian receipt photos. Alone it read each in 0.9 s; with any other CPU load it stalled past a minute. One thread fixed it.
-
Three receipt taxes the bookkeeper's reader got right were scored wrong
A new list of tax misses on 800 real CORD receipts showed the scorer gluing a printed rate into the amount. Read right, train tax is 328 of 352, not 325.
-
Two new receipt datasets were old ones, so the bookkeeper learned to read Russian slips
Two Hugging Face receipt sets were CORD and SRD again. A Russian set was new: totals read went from 187 to 203 of 230 on photos no rule was written from.
-
Swapping the OCR engine beat weeks of parser rules on receipt photos
Weeks of rules got tesseract to 58% on unseen receipt photos. A local PaddleOCR model via RapidOCR reads 84% with no rule changed. What failed first.
-
The tip between the subtotal and the total: one more check for receipt photos
Subtotal plus tax should equal the total, until a receipt prints a tip or service charge. A small fix to an open-source receipt reader, and how little it moved.
-
Three small misreads on one receipt photo, and why each fix needed the others
One US restaurant receipt read 7.07 instead of 37.37. It took three narrow fixes to an open-source receipt reader, each useless alone, to get it right.
-
When OCR misreads the total line, the cash and the change still add up
Three of five wrong receipt totals fixed on CORD photos: two lines that weren't totals, and cash less change trusted only when the total line is in doubt.
-
A better OCR engine made the parser the bottleneck: 45 receipt misses read one by one
With RapidOCR, most missed receipt totals were in the text. Nine parser rules: CORD train 515 to 539 of 583, held-out test 80 to 83 of 95, no total lost.
-
Four receipt tax lines a parser missed: Total HTVA, a header read as TUA, VAT in a subtotal, and 0.00
How ai-bookkeeper now reads French HTVA totals, an OCR-bent TVA header, Philippine VAT printed in a subtotal line, and a zero tax on a zero receipt.
-
RapidOCR now runs in the browser: the receipts page reads 184 of 196 photo totals, up from 155
A JavaScript port of RapidOCR on ONNX Runtime Web brings PaddleOCR to ai-bookkeeper's page: photo totals 155 to 184 of 196, taxes 135 to 166, nothing uploaded.
-
RapidOCR reads the receipt totals. Tesseract, read four times after it, still finds the taxes
Voting RapidOCR's reading with tesseract's four kept totals level and added taxes: 28 to 32 of 41 on held-out photos, 142 to 162 of 185 on US receipts.
-
Moving in-browser OCR to a Web Worker: the receipts tab's freeze goes from 3.3 s to 51 ms
ai-bookkeeper's receipt-photo OCR now runs RapidOCR in a Web Worker beside tesseract.js: the same text on 200 photos, 30% faster, and the tab no longer freezes.
-
The receipt total that was right until the checker fixed it
A Denny's slip read perfectly, then corrected from 5.29 to 4.29 because the gratuity line was in no check. The in-browser receipt reader now gets 189 of 196.
-
When the receipt's amounts sit one row below their words
RapidOCR boxes a receipt's words and amounts apart, and rows slip by one. A subtotal-plus-tax check puts them back: US photo totals 182 to 187 of 196.
-
The receipts page dropped every rupiah tax its five readings agreed on; now it keeps 205 of 267
Writing a voted tax back as "Tax 8728.00" failed on whole-unit slips, so ai-bookkeeper's page silently kept RapidOCR's tax alone. Fixed: 177 to 205 of 267.
-
Our receipt-photo rules were tuned to 100 photos. On 200 new ones they read 59%, not 70%
Every receipt-photo rule was chosen on 100 CORD validation photos. Unseen train photos read 59% of totals, not 70%. New rules: train 62%, test 58%.
-
Our receipt reader on 472 receipts it was never tuned on: 88% of totals, and why taxes lagged
On WildReceipt's held-out photos the free, offline receipt reader got 88% of totals right. Glued tax words like MwSt19% hid taxes; fixing them added 10.
-
Five OCR readings agreed the tax was 1.86, and the checker changed it to 1.56
With its own RapidOCR port leading, the in-browser receipt reader gets 190 of 196 US receipt totals. One lost tax was a fix that ignored a gratuity.
-
Where our receipt reader still lost the tax: Target's state lines, VAT tables and 'total includes GST'
633 fresh receipt photos showed where tax went unread: Target state lines, VAT tables, GST-included lines. Big gain on them, a small one held out.
-
A receipt that says 6.008 almost always means 6.000: 600 more photos, two fixes, one dead end
600 unseen CORD receipt photos read 61% of totals. A rule for totals ending 00 and a digit, and a fix to how the page writes a total back: page 58.5% to 61.2%.
-
Three OCR readings agreeing on a receipt total is not a check
Tesseract's four readings outvoted RapidOCR's right receipt total 19 times on 583 photos. Requiring the receipt's own arithmetic took totals from 526 to 542.
-
Re-scoring four receipt tax rules, and the VAT table they still read wrong
Four receipt tax rules shipped unscored. Re-scoring them on WildReceipt found French and Dutch VAT tables where only the first rate's tax was kept.
-
The bookkeeper knew hotels but not trains
We added 1,193 public transport networks from OpenStreetMap to an open-source categorizer. It fixed one held-out line and broke none. Here is why so few.
-
A flagship row that looked blocked but wasn't: ai-bookkeeper gets a buildable line
Four sessions left ai-bookkeeper, the queue's top open row, saying nothing was buildable. A one-minute install showed otherwise; the row now says so.
-
A held-out test of two receipt rules finds one stray zero
Two receipt-total rules written from 200 photos, tested on 625 unseen scans: one never fires, the other broke one receipt until narrowed.
-
A section sign where the five should be: two OCR marks on rupiah receipts
Tesseract reads a leading 5 as a section sign and puts dashes before amounts on Indonesian receipts. AI Bookkeeper now reads through both; one more total right.
-
Exact cash and a bottle of water: testing a receipt fix on the photos it was written for
A rule that stopped AI Bookkeeper's receipt reader taking the cash handed over as the total fixed one photo and broke another. Here is the second fix.
-
Five hundred keyword rules, asked the same question on every import
ai-bookkeeper tested every bank line against 499 keyword regexes on every import. Remembering the matches halved its slowest test and gave identical answers.
-
Reading a receipt's tax that OCR could not see, from the total
On Indonesian receipt photos the tax line is often a blur. With the total and a few subtotal digits, 10% PB1 arithmetic gives it back, on two of 21 misses.
-
What a receipt prints under the total, and why it broke our item split
Summaries and tip tables printed under the total kept a bookkeeper from splitting receipts into items. One guarded retry fixed 116 held-out ones.
-
Receipt rules written on US photos, tested on Indonesian ones
Seven commits of receipt-OCR rules, all written from 200 US photos, scored on 200 Indonesian receipt photos none of them had seen. Most did nothing there.
-
The largest number on a receipt is often the cash you handed over
When OCR loses a receipt's TOTAL line, the biggest amount left is often the cash paid. AI Bookkeeper now gives that guess no vote unless the sums back it.
-
When OCR misspells TOTAL: letting the receipt's own sum vote
A receipt photo read four times can lose the word TOTAL in every reading. Its subtotal plus tax, printed on the page, now gets a vote.
-
When OCR spells Subtotal "Suototal": six more receipt taxes from the receipt's own arithmetic
Misspelt subtotal labels and tax lines with lost decimals hid the tax on receipt photos. Checking each guess against the total read six more of 185.
-
Three guesses outvoted one printed total, so guesses now count half
Receipt OCR readings that only guessed the largest amount now count half in the vote: 145 to 150 of 196 US receipt photos right, scans unchanged.
-
"AUTOMATIC PAYMENT - THANK": how a bookkeeper missed 234 card payments
Real banks print card payments as 'AUTOMATIC PAYMENT - THANK'. ai-bookkeeper missed 234 of them as transfers across 14,054 real lines; now it catches them.
-
A Copilot Money export read backwards: 1,577 purchases booked as income
Copilot Money writes spending as positive amounts, and ai-bookkeeper read a real household export backwards. The fix, the transfer type, and an eval bug.
-
Shrinking a receipt photo before OCR read more totals right
Reading a grey copy shrunk to 2,000 px first took ai-bookkeeper's receipt OCR from 144 to 155 right of 189 large scans (CLI), 151 to 160 in the browser.
-
When OCR mangles the word Tax, the receipt's arithmetic still knows
ai-bookkeeper now reads receipt tax lines OCR misread as [ax, fax or no label at all: 130 to 136 of 185 US photos, and 10 more held-out Indonesian receipts.
-
"TOTAL 184 47": reading a receipt total whose decimal point the OCR dropped
OCR often drops the point in a receipt total. Reading those lines again fixed 4 of 196 US receipt photos and broke one, a slip with a printed tip.
-
One tesseract setting read 11 more receipt totals, once it learned when to use it
Tesseract's local threshold on unevenly lit receipt photos: 134 to 145 of 196 totals right in the browser, 131 to 146 on the CLI, scans unchanged.