We have released the AI image counting system 'ai-Numbers' that instantly determines the number of sheet materials and thin products. This product analyzes electronic circuit boards with complex edges and sheet products with significant cross-sectional variations from the side at high resolution using AI, automatically determining excess or shortage in as little as 0.3 seconds (※). Furthermore, our unique learning method eliminates the need for threshold settings, fundamentally reducing counting errors while shortening setup time, thereby alleviating the workload in manufacturing sites. (※) Depends on the field of view and counting conditions. Conventional non-contact counting machines extracted one-dimensional brightness from images and counted changes in that brightness. As a result, they could not accurately count due to changes in the state of the edges or because items in transparent bags were partially visible. ai-Numbers uses AI to extract two-dimensional features such as color, patterns, and textures of the edges. This allows for accurate counting of electronic circuit boards with complex edges and sheet products with significant cross-sectional variations. With a maximum resolution of 5,400 pixels capturing the edges, it can count thin sheets of 0.05mm, which was impossible with conventional image counting machines.

In November last year, we released the packaged version of 'NumberVisionAPI' as a new product called 'NumberVision', a license plate recognition system! ■ NumberVision instantly recognizes and digitizes vehicle license plates. This enables various applications such as parking management, enhanced facility security, customer management, and traffic statistics. It can recognize plates with high accuracy even in dark places or bad weather, and since it is a one-time purchase, there are no running costs! It can be flexibly utilized according to the customer's environment, allowing for license plate recognition in diverse situations. ■ Use Cases Monitoring vehicles at factories and construction sites, managing employee parking lots, monitoring unauthorized parking during long stays or outside business hours, managing vehicle entry at hotels and car dealerships, streamlining maintenance reservations, displaying visitors at small businesses and private practices, confirming vehicles during closing hours, simplifying reception tasks in security offices or guardhouses, managing automatic gates in apartments, conducting traffic volume surveys, collecting statistics on visitors at roadside stations, measuring dwell time, reducing costs for coin parking, and managing vehicles at waste disposal facilities, among others. From monitoring vehicles at factories and construction sites to managing employee parking lots, it can be used in a variety of scenarios.

We have released the NumberVision API, a license plate recognition API specialized in vehicle number plate reading performance! It instantly recognizes vehicle numbers from captured images and digitizes the data using AI technology. Software and web service developers can easily incorporate it into their systems. 【Supports license plate recognition in various scenarios】 - Management and operation of parking lots - Entry and exit management for factories and buildings - Hotel parking lots - Car dealerships - Commercial facilities - Highways and general roads 【Three features of the NumberVision API】 ■ Recognition rate of 99.9% With multi-plate reading for car sales, it can simultaneously read multiple license plates from a single image. ■ Usable in various locations It enables license plate recognition in scenarios that were previously considered difficult, allowing operation in places with vehicle entry and exit other than parking lots, both indoors and outdoors. ■ No initial setup costs It operates on a monthly subscription basis, allowing you to use only what you need when you need it, without any initial costs. You can experience it on the product page! Please proceed from the related link.

We have released the inspection machine "RakuiroChecker," which determines the color order of wires the moment the wire harness is presented. It prevents defects from leaking due to human error. 【Features】 ■ If the housing and wires are visible on the screen, it will determine pass or fail regardless of their position. Moreover, the judgment time is 0.3 seconds. You can perform inspections one after another while holding a bundle of wire harnesses in one hand. ■ Using a unique algorithm, it learns the front and back of the housing and simultaneously judges both sides. There is no need to worry about the orientation of the housing when presenting the harness to the camera. ■ It detects changes in ambient brightness in real-time and automatically adjusts the color judgment settings. Therefore, it is less affected by the surrounding brightness environment and can be used anywhere. As a next-generation miswiring inspection tool, RakuiroChecker is equipped with many new features such as high-speed image processing, automatic judgment of the housing's front and back, and easy operation by simply presenting it. As the variety of small-lot products increases, it reduces the burden on inspectors and prevents miswiring from leaking. We hope you find it useful.