The NuMicro M480 series is based on the Arm kernel kernel of Cortex M4, with digital signal processing instruction (DSP) and floating point number computing unit (FPU). The main frequency is up to 192 MHz, which greatly improves the execution efficiency of machine learning neural network algorithm. This system can support 320x240 image resolution input, the number of frames up to 30 FPS, and support i80 or SPI interface LCD screen to real-time display images. Up to 8 UART can be connected to many different wireless pass-through modules at the same time to complete remote data transmission.

It features a 192 MHz cortex-m4, DSP and FPU, which can complete digital identification in 0.5 seconds.

Can support QVGA (320x240) image capture and CMOS camera interface support CCIR601/ CCIR656 transmission; Not only that, but also support Arm Pelion, AWS, aliyun and other cloud services.

Image recognition, for example, has long been a hot topic: image information is captured by image sensors and then analyzed and identified by processors, which convert image information into text. Examples commonly heard in daily life are Optical Character Recognition (OCR), in which an article is scanned, its text is identified, and then converted into editable text files in a computer. In the past, such actions were done by computers because of the strong computing power required. For image recognition, new don gives you another option, we use microcontrollers instead of computers for simple digital identification.

We use the Nuvoton NuMicro M480 series microcontroller and machine Deep Learning algorithm to capture the Numbers on the plate of traditional water meter and electricity meter through CMOS sensor, identify the exact 0-9 Numbers, and then transmit the data to the back end through nb-iot, LoRa or wi-fi, so as to complete the function of remote meter reading. The reason why remote meter reading is important is to save a lot of labor costs. In the past, the calculation of water and electricity bills required a person to go door to door to record the figures on the disk, and then unified distribution of bills. Now, as long as a digital identification device dominated by the new tang NuMicro M480 series microcontroller is added to the traditional disk, all data can be sent back to the back-end for subsequent processing within minutes, saving the cost of replacing the traditional table with a smart meter.

More details, pls visit: https://www.indasina.com/nuvoton-microcontrollers_c29