High-speed data collection device and method based on FPGA

Abstract

The invention discloses a high-speed data collection device and method based on an FPGA. The system comprises a binocular camera, an IMU, an FPGA platform and an internet access module, the binocularcamera and the IMU are connected to the internet access module through the FPGA platform, and tracking signals of two image sensors of the binocular camera are synchronously triggered by the FPGA platform and synchronously output collected real-time images to the FPGA platform; the IMU and the two image sensors achieve hardware synchronization through the FPGA platform; the FPGA platform comprisesan FIFO and a binocular IMU transmission IP core. Wherein the FIFO is connected with the IMU and the two image sensors, stores real-time binocular images and IMU data at the same time, is connected with the binocular IMU transmission IP core, and transmits synchronous binocular images and IMU data to the internet access module through the binocular IMU transmission IP core. According to the invention, binocular images and IMU data can be synchronously transmitted in a low-delay manner, and the method is suitable for the fields of visual navigation positioning, industrial detection and recognition.

Description

technical field [0001] The invention relates to the technical field of machine vision, in particular to an FPGA-based high-speed data acquisition device and method. Background technique [0002] Binocular image data is widely used in the field of visual navigation and positioning, such as: UAV visual inertial navigation positioning system, automatic driving navigation obstacle avoidance system, AR, VR, indoor robot and warehouse AGV positioning and navigation system, etc.; binocular image data is also Widely used in the field of recognition, such as: 3D face recognition, object perception tracking. At the same time, binocular image data is also used in measurement fields, such as depth measurement, ADAS and industrial screening. [0003] At present, some binocular image acquisition solutions use STM32 as the main control, it is difficult to process the data collected by two cameras at the same time, and the heat is relatively serious; some synchronization errors between bin...

AI Technical Summary

Problems solved by technology

[0003] At present, some binocular image acquisition solutions use STM32 as the main control, it is difficult to process the data collected by two cameras at the same time, and the heat is relatively serious; some synchronization errors between binocular image data and IMU data reach tens of milliseconds. basically unusable

Some binocular image data transmission delay reaches more than 100 milliseconds, and cannot be transmitted in real time. It is basically unusable in the field of drones and automatic driving.

Some binocular image data transmission bandwidth is not enough, and the packet loss is serious in the high frame rate mode, which cannot achieve high frame rate real-time detection, and basically cannot be applied in the field of industrial detection

Images