| CPC G06T 3/4092 (2013.01) [G06T 5/20 (2013.01); G06T 5/50 (2013.01); G06T 5/92 (2024.01); G06T 5/94 (2024.01); G06T 7/0002 (2013.01); G06V 10/751 (2022.01)] | 19 Claims |

|
1. A testing system for an image processing algorithm, comprising:
a control unit, configured to provide an original image and a parameter setting value;
an image processing device, connected with the control unit and configured to obtain the original image and the parameter setting value; an image processing hardware, connected with the image processing device, and configured to be driven by the image processing device to perform a first image processing procedure to the original image in accordance with the parameter setting value and generate a hardware-processed image; and
a testing device, connected with the control unit and the image processing hardware, configured to obtain the hardware-processed image from the image processing hardware, and to obtain the original image and the parameter setting value from the control unit or the image processing hardware;
wherein, the testing device is configured to execute a simulation software and a testing software, the simulation software is executed to perform a second image processing procedure to the original image in accordance with the parameter setting value and generate a software-processed image, the testing software is executed to perform a comparing procedure to compare the hardware-processed image with the software-processed image and generate a comparing result, wherein the second image procedure is identical to the first image procedure, and the comparing result shows a pixel difference between the hardware-processed image and the software-processed image,
wherein the first image procedure and the second image procedure comprise a plurality of algorithms whose execution are scheduled in a sequential order, wherein the testing system creates a testing point (TP) whenever one of the plurality of algorithms is executed completely, and generates and compares the hardware-processed image with the software-processed image at every testing point,
wherein the plurality of algorithms comprise de-bayering procedures, gamma correction procedures, scaling procedures, or dithering procedures;
wherein, when the testing system determines that at least one of the pixel differences that is greater than a tolerance threshold exists between the hardware- processed image and the software-processed image at a certain one of the testing points, the testing system adjusts the algorithm being used or the parameter setting value being used before the certain one of the testing points.
|