Through the establishment of a deep learning image recognition system for city gas meter guidelines (metric numbers), the automatic recognition of image numbers from the meter guide photos is applied to minimize unnecessary manual work on meter reading data and work automation to improve city gas business and corporate competitiveness. We try to improve.
Gasmeter guideline recognition range
· Analog Gasmeter - integer numbers 4 to 6 · Digital Gasmeter - integer numbers 4 to 6 - When uploading a photo, a digital Gasmeter photo is required.
Recognizing gasmeter guidelines
· Provide a deep network that has been trained with gasmeter guidelines · Recognize meter instructions from meter photos through deep learning engine
Interlocking support
· Receive gasmeter photo · Interworking with the gasmeter guide value business system recognized by deep learning
Providing deep network of the optimal gasmeter
· Learning about various gasmeter models · Learning about gasmeter photos taken in various environments · Minimize wrong detection by applying optimal threshold
Fast recognition speed
· Fast recognition speed using GPU: 0.05 seconds · GPU: Graphics Processing Unit
Gasmeter recognition flow chart
Gasmeter recognition batch process
Real-time process of gasmeter recognition
Gasmeter recognition process result
OS
· Windows 10 Professional 64 bit or higher (when configuring high-end PC Workstation) · Windows 2012 Server Standard 64 bit or higher
Server specification
OS · Intel Xeon 2.3Ghz 4-Core or higher · Memory 16 GB or higher · Memory 16 GB or higher · HDD 300 GB or higher GPU version · Intel Xeon 2.3 Ghz 4-Core or higher · Memory 16 GB or higher · HDD 300 GB or higher · GPU : NVIDIA Quadro-P5000 or higher (GPU when configuring high-end PC Workstation : GTX 1080Ti 11g)