COLONYCOUNT AI
AI-powered bacterial colony counter for microbiologists. Upload your plate image — get an accurate CFU count in seconds. No manual counting. No eye strain.
How it works
No calibration. No setup. No manual counting. Just upload and get your result.
Photograph your agar plate under consistent lighting and upload the image — JPEG or PNG. No special equipment required beyond a standard camera or phone.
Computer vision scans the plate, identifies individual colonies based on size, shape, and contrast, and distinguishes colonies from artifacts and background variation.
Receive a total colony count with a visual overlay marking each detected colony. Download the annotated image and count data for your lab records.
Features
Colony counting is tedious, time-consuming, and subject to inter-observer variability. ColonyCount AI removes all three problems.
Computer vision trained on microbiological plate images. Detects colonies across varying sizes, densities, and media types with consistent accuracy.
Manual plate counting takes minutes per plate and accumulates fatigue errors over a session. ColonyCount AI returns results in seconds regardless of colony density.
Eliminates inter-observer variability — the same plate image returns the same count every time. Critical for experiments where CFU consistency affects downstream analysis.
Download the plate image with each detected colony marked. Includes the colony count and timestamp — ready to attach to your experiment record or ELN entry.
Works with TSA, LB, MacConkey, blood agar, and most common microbiological growth media. Handles circular and square plates in standard laboratory photography conditions.
Export your colony count directly into a BenchVoice lab notebook entry. Your plate count becomes part of your searchable experiment record automatically.
Use cases
Anyone who counts colonies manually is a candidate for a faster, more consistent approach.
Growth curves, antimicrobial susceptibility, kill assays — any experiment requiring accurate CFU counts.
Total plate counts, yeast and mould counts, and indicator organism quantification for food safety and QC.
Reduce time spent on manual counting during high-throughput experiments. More plates, less fatigue.
Environmental monitoring and bioburden testing where colony count accuracy and documentation matter.
Standardise student colony counting results and use AI-annotated images for instructional comparison.
Semi-quantitative culture counts and urine culture colony enumeration where speed and consistency are critical.
FAQ
ColonyCount AI performs comparably to experienced manual counters on standard plates with well-separated colonies. Accuracy is highest on plates with 30-300 colonies — the standard countable range. Heavily confluent plates or plates with unusual morphology may require manual review of the annotated output.
ColonyCount AI works with standard agar plates including TSA, LB, MacConkey, blood agar, R2A, and most common microbiological growth media. It supports JPEG and PNG image formats. Best results come from photographs taken with consistent overhead lighting and a plain contrasting background.
Yes. ColonyCount AI is free to use. Upload your plate image and get a colony count with no account required. For high-volume use or API access, contact JoeLuT AI about research or institutional plans.
Phone photos work well provided the lighting is consistent and the plate is photographed straight-on without shadows or glare. A simple lightbox or consistent overhead lighting setup significantly improves detection accuracy.
Heavily confluent plates — where colonies are touching or overlapping throughout — are challenging for both AI and manual counting. ColonyCount AI will flag high-density regions and provide an estimated count, but plates with TNTC (too numerous to count) density should be diluted and re-plated for accurate quantification.
COLONYCOUNT AI
Free. No account needed. Results in seconds.