“The leading semiconductor components distributor dedicated to the Asia-Pacific market, Dalian General Holdings, announced that its subsidiary Quanding has launched a multi-finger gesture recognition solution based on PixArt PAC7640LT + PAG7661QN.

“

On May 11, 2021, Dalian General Holdings, a leading semiconductor component distributor dedicated to the Asia-Pacific market, announced that its subsidiary Quanding launched a multi-finger gesture recognition solution based on PixArt (PixArt) PAC7640LT + PAG7661QN.

Figure 1-Scenario application diagram of the multi-finger gesture recognition solution based on PixArt products launched by Dalian Daquanding

Gesture recognition technology has emerged to make communication between people and smart devices more natural and comfortable. In recent years, with the rapid development of smart terminals, the high popularity of AR/VR, and the acceleration of the Internet of Everything, gesture recognition has been sought after by major manufacturers as a more humanized interaction method. Dalian Daquan Ding launched a multi-finger gesture recognition solution based on PixArt PAC7640LT + PAG7661QN, which can support single object/multiple object detection, and provide more accurate and reliable gestures for various applications including business, IoT and games product.

Figure 2 – The Display board of the multi-finger gesture recognition solution based on PixArt products launched by Dalian Daquanding

Founded in July 1998, PixArt is mainly dedicated to the design, development, production and sales of CMOS image sensors and other application ICs. PixArt has rich experience in analog IC design, image sensor IC design and image processing IC design, and is now quite influential among global CMOS image sensor manufacturers.

This solution combines PAC7640LT (Sensor) + PAG7661QN (DSP) + LST0-227 (Lens) with an independent ISP SoC, with complete algorithm support, providing a comprehensive solution for product development. This solution supports multiple gesture recognition: stationary fist or 1-5 fingers extended gesture, 1-5 fingers extended to move towards the sensor, two-finger pinch and unique thumb mode, etc.

Figure 3 – Block diagram of the multi-finger gesture recognition solution based on PixArt products launched by Dalian Daquanding

Among them, the PixArt PAC7640LT is an enhanced global shutter image sensor with a package size of 3.45mm x 3.45mm x 0.76mm (CSP). It has 400 x 400 resolution, 120 frames per second (fps) image capture capability, and the sensor comes with an integrated image sensor ISP. In addition, the PAC7640LT allows synchronization of multiple cameras, supports 3D type applications, and can control all ISP functions through register settings via I2C.

PAG7661QN itself is a programmable digital signal processor (DSP), mainly through MIPI to convert DATA/CLOCK/FSYNC to USB/I2C/SPI, through I2C or SPI or USB1.1 can immediately read the predefined As a result of gestures, MCU processing resources are greatly saved.

Core technical advantages:

• Based on infrared image analysis technology, the hardware supports hand shape recognition algorithm;
• Built-in hardware algorithm, fast response;
• Support real-time output of palm and finger signals to increase application flexibility;
• The predefined gesture results can be read through I2C, SPI or USB1.1, which greatly saves MCU computing resources;
• Miniaturized design;
• Free hardware reference design, software reference code.

Program Specifications:

• Operating Distance: 5~40cm;
• Communication interface:
• I2C, SPI and USB1.1 for Gesture result output;
• USB 1.1 for Image mode (50×50);
• 5V power supply;
• 2 LEDs;
• FOV(horizontal) = 77deg;
• 4L PCB, dimensions: 55.5mm x 10mm.