High-speed lossless ultrahigh-definition industrial visual detection method and system

Abstract

The invention relates to a high-speed lossless ultrahigh-definition industrial visual detection method and system. The method comprises the following steps that: at an offline configuration stage, a smart camera unit consisting of an ultrahigh-definition image sensor, an ISP (Internet Service Provider) processing module, a visual detection module, a hardware clock synchronizing module and the like remotely transmits high-resolution RAW data and RGB (Red Green Blue) image data to a computer through an optical fiber to generate a detection algorithm and a control logic; and at a high-speed online running stage, a visual detection module performs detection operation, marks a feature image, remotely transmits the feature image to a high-speed recording and playing unit through the optical fiber to perform recording, and asynchronously displays the feature image on the computer. Through adoption of the high-speed lossless ultrahigh-definition industrial visual detection method and system, the original RAW data and the RGB image data are detected directly and visually through an FPGA (Field Programmable Gate Array) module, and remotely recorded synchronously through a multichannel optical fiber, so that real-time performance of the high-speed ultrahigh-definition visual detection system can be effectively ensured.

Description

Technical field [0001] The invention relates to a high-speed non-destructive ultra-high-definition industrial vision detection method and system, which belongs to the field of industrial cameras and industrial machine vision computer detection, especially the high-speed ultra-high-definition machine vision detection field. Background technique [0002] With the development of advanced manufacturing, high-precision manufacturing has increasingly higher requirements for computer machine vision inspection technology. Therefore, whether the speed of large-format visual online inspection can meet the needs of rapid production, on the one hand, industrial cameras are required to have higher performance, and on the other hand, the balance between computing speed and resource consumption must be maintained at a certain level. At present, the machine vision inspection system mainly has two modes. One is the independent operation of the smart camera, which is mainly used for simple workpie...

AI Technical Summary

Problems solved by technology

At present, the machine vision inspection system mainly has two modes. One is that the smart camera operates independently, which is mainly used for simple workpiece positioning or defect detection, and uses small-format images; the other is that the industrial camera transmits the collected images to the computer through the network. , the computer performs visual detection processing and control output. With the increasing application of ultra-high-definition images in production, certain requirements are put forward for detection speed and synchronization. Therefore, the inherent large image data of this method Problems such as parallel processing, high-bandwidth image transmission bottlenecks, and huge consumption of computing resources are difficult to solve

Images