Product Introduction
Product Description
AidLux is a fusion operating system built on ARM hardware, based on an innovative cross-Android/HarmonyOS and Linux system environment. SmartVision is an intelligent vision application platform that includes an application development framework for intelligent vision computing terminal devices, intelligent vision terminal device configuration management, application runtime services, and lifecycle management services. Running within the AidLux system, it leverages the AI application runtime SDK provided by the AidLux platform to achieve defect detection of target objects, providing users with a comprehensive defect detection solution.
Product main features
- Supports 12Tops computing power and can adapt to AI deep learning algorithms in various scenarios
- Built-in mature algorithms for defect detection, classification, object recognition, and character recognition
- Support secondary development and provide full-process AI tool chain software
- Supports access to peripherals such as monitors and mice for easy maintenance
- The product has rich interfaces, supports IO signal input and output, and is suitable for various industrial detection scenarios.
Product appearance introduction
The overall structure of the product is small and compact, with high flexibility. The product shell is made of aluminum alloy and has a surface treatment process of sandblasting and then black oxidation, which not only increases the texture of the product, but also takes into account the heat dissipation of the product; it is partially equipped with orange decorative strips. It adds bright colors to the product, making it calm and atmospheric, but not monotonous. As shown below:
Product appearance and installation dimensions
The maximum overall dimensions of the device are length * width * height =110 * 81.2 * 82mm.There are two sets of mounting holes on the side and back respectively. They are equipped with M4 threaded holes and the hole spacing is 26 * 26mm. As shown below:
Product components and interface description
| number | name | describe |
|---|---|---|
| 1 | Mounting screw holes | 4 M4 mounting screw holes, effective thread depth 8mm, hole spacing: 26*26mm |
| 2 | 10PIN interface | See Figure-Power supply and I/O interface definition |
| 3 | 20PIN interface | See Figure-Power supply and I/O interface definition |
| 4 | LED indicator | Network port status indication |
| 5 | Ix Industrial lan port | Industrial IX Gigabit Ethernet network port |
| 6 | USB dual layer interface | 2-way USB 2.0 Type A |
| 7 | USB side stand interface | 1 USB 2.0 Type A |
| 8 | HDMI side stand interface | HDMI mini display output |
| 9 | Type-C interface | 1 channel Type-C OTG |
| 10 | Power interface | DC 12V IN |
| 11 | LED indicator | Machine status indication |
| 12 | Micro SD interface | Expanded high-speed TF card storage interface |
| 13 | Power button | Press and hold for 15s to restart |
| 14 | C-mount lens | Lens focal length 4-20mm optional |
Product interface introduction
Power supply and IO interface definition
10Pin and 20Pin pin definition
| Pin | Signal | I/O signal source | illustrate |
|---|---|---|---|
| 10-1 | CAN0 H | CAN0 | |
| 10-2 | CAN0_L | CAN0 | |
| 10-3 | GND | GND | |
| 10-4 | CAN1_H | CAN1 | |
| 10-5 | CAN1_L | CAN1 | |
| 10-6 | GND | GND | |
| 10-7 | RS485_M | RS485 | |
| 10-8 | RS485_P | RS485 | |
| 10-9 | GND | GND | |
| 10-10 | 5V | 5V output | |
| 20-1 | GPIO_12 | SPI_MISO | SPI master device output interface (slave device input interface) |
| 20-2 | GPIO_13 | SPI_MOSI | SPI master device input interface (slave device output interface) |
| 20-3 | GPIO_14 | SPI_SCLK | SPI clock interface |
| 20-4 | GPIO_15 | SPI_CS_N | Chip select |
| 20-5 | GND | Ground | |
| 20-6 | GPIO_22 | SM_DBG_UART_T X | Debug UART send |
| 20-7 | GPIO_23 | SM_DBG_UART_R X | Debug UART receive |
| 20-8 | OUTPUT_GND | Optocoupler output ground | |
| 20-9 | GPIO_61 | OUTPUT_1P | Optocoupler output 1 |
| 20-10 | GPIO_6 | OUTPUT_2P | Optocoupler output 2 |
| 20-11 | GPIO_52 | INPUT_2P | Optocoupler input 2 |
| 20-12 | GPIO_81 | INPUT_1P | Optocoupler input 1 |
| 20-13 | INPUT_GND | Optocoupler input ground | |
| 20-14 | GPIO_98 | I2S_FS | I2S master clock_MCLK |
| 20-15 | GPIO_100 | I2S_SCLK | I2S bit clock_SCK |
| 20-16 | GPIO_97 | I2S_DIN | I2S field selection_WS |
| 20-17 | GPIO_96 | I2S_DOUT | I2S data channel_DATA |
| 20-18 | GND | Ground | |
| 20-19 | UART0_TXD | Module serial port_TX | |
| 20-20 | UART0_RXD | Module serial port_RX |
Product software architecture
The overall product architecture is shown in the diagram below:
- AidLite SDK: A component that helps developers efficiently develop neural network model inference functions on the AidLux fusion system platform. AidLite SDK supports various inference model formats, including open-source formats and proprietary formats from AI chip manufacturers (such as TFLite, Qualcomm DLC, Rockchip RKNN, etc.), and supports extreme acceleration of model inference on various heterogeneous computing units (such as CPU, GPU, DSP, NPU, etc.).
- SmartVision SDK: A component that helps intelligent camera application developers with the development of image acquisition-related functions and data integration-related functionalities.
- SmartVision Engine (abbreviated as SVE): An intelligent camera management and basic service platform. It is used to view the status of intelligent cameras, configure imaging parameters, and provide deployment and lifecycle management capabilities for standard intelligent camera applications (SmartVision App).
- SmartVision App: Developers can use SmartVision SDK to develop smart camera standard applications that can be integrated and managed by SVE.
Basic Workflow
The basic workflow of an AI application and the key SDK interfaces involved in the process are shown in the diagram below:
