Tofu Counting

csdetb-icono

Customer

One of Japan’s leading Fried tofu manufacturer.

Project Details

  • Duration: 1 Month +

Technologies:

  • Deep Learning
  • OpenCV
  • Python
  • PyQt
  • MongoDb
  • Edge computing device
problem-img

Problem

  • The tofu manufacturer faces challenges in accurately counting the fried tofus going into a single packaging.
  • Manual counting is prone to errors due to fast-moving lines.
  • It also fails to provide real-time insights and statistics on the efficiency of each line. This limits the benefits of real-time production monitoring.

 

solution-img

Solution

We developed a desktop application that detects and counts fried tofu using a segmentation model.

  • The count is updated in real time and displayed on-screen for each package pass.
  • Daily reports can be easily exported and analyzed by improving the workflow of food production automation.
  • The solution is designed for quick integration into existing food packaging automation lines.
  • Built to support food processing automation. It increases speed, accuracy, and operational clarity.
  • Helps with packaging line optimization by identifying trends in discounts or drops in output.

99.56%

Accuracy

< 1s

Detection & Counting Speed

Implementation-Challenges

Implementation Challenges

  • Achieving accurate segmentation and counting on fast-moving tofu pieces.
  • Ensuring consistent results under different lighting and frying conditions.
  • Integrating with existing automated food packaging systems without halting operations.
  • Providing exportable data that aligns with production reports.
  • Designing the UI for non-technical plant operators.
Key-Features

Key Features

  • Real-time tofu detection and counting display.
  • Compatible with high-speed production lines. 
  • Easy daily report export for batch tracking.
  • Seamless integration into food packaging automation environments.
  • Supports operational scalability across multiple lines.
  • Improves overall efficiency in food processing automation.

Insights on innovation

Stay updated with the trending and most impactful tech insights. Check out the expert analyses, real-world applications, and forward-thinking ideas that shape the future of AI Computer Vision and innovation.

June 19, 2026 - 4 minutes to read

The Clock You Cannot See: How to Stop Demurrage Before Midnight

It is 11:40 on a Thursday night. The yard is dark, the gate is shut, and Suresh, the duty manager, has gone home. On a screen in an empty control room, three container records tick past midnight. At 12:00, their free time ends. The clock starts. A few rupees an hour, then a few more, […]

Read More

Ruchir Kakkad

CEO & Co-founder

June 18, 2026 - 4 minutes to read

Why Your Yard Is Never as Full as the System Says

The vessel is twelve hours out and Anjali, the yard manager, is staring at a screen that says the yard is 92 percent full. She has a stack of boxes inbound and nowhere obvious to put them. So she does what experienced managers do under pressure: she books overflow space at the depot down the […]

Read More

Ruchir Kakkad

CEO & Co-founder

June 17, 2026 - 5 minutes to read

Inside One Gate Pass: What Container Gate Automation Reads in 8 Seconds

It is 9:14 on a wet Tuesday and Ramesh has been in the gate booth since six. The rain is sideways now. A line of trucks idles back toward the road, engines warm, drivers impatient. Ramesh leans out, squints at the side of a container, and copies eleven characters onto a damp sheet. He has […]

Read More

Ruchir Kakkad

CEO & Co-founder

Whatsapp Img