Verification platform and verification method

Abstract

The invention provides a verification platform, comprising a video incentive generation module, the video excitation generation module receives the input video file, and generates a RAW RGB input file; a verification module, the verification module is connected to the video excitation generation module by communication, receives And process the RAW RGB input file, and output the image to be compared and the reference image; the display module is connected to the verification module by communication, and receives and visualizes the image to be compared and the reference image sent by the verification module The image is for testers to observe the image processing effect of different configurations, which solves the problem that the traditional verification platform cannot satisfy the verification of the image processing module, and develops a verification platform specially for the image processing module, which can simulate the stimulus input of real video stream data at the same time. It allows testers to test and compare images from different perspectives in a set of verification platforms. The invention also provides a verification method based on the verification platform.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a verification platform and a verification method. Background technique [0002] The image processing module (Image Signal Processing, ISP) is used to process the RAW RGB format image signal transmitted from the sensor chip. Since hardware has a faster processing speed than software, hardware such as integrated circuits is widely used to realize high-speed and high-performance processing of original images. The YUV series image format is widely used in the field of AI intelligent recognition because of its advantages of convenient image compression, high-speed transmission and processing, and most importantly, it can retain the most important Y component (brightness) grayscale information in AI recognition. At the same time, the ISP module will generate RGB images. For example, the AI ​​module uses bitmap algorithms (Bitmap, BMP) as input data to train the neural networ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a verification platform and verification method to solve the problem that the traditional UVM verification platform cannot satisfy the verification of image processing modules, to simulate the excitation input of real video stream data, and to allow testers to verify in a set of visual processing Test and compare image processing results from different perspectives on the platform

Images