Detection device and method for measuring xy-directional displacement of target object in three-dimensional space using monocular camera

Abstract

This invention relates to a detection device and method for capturing planar images in three-dimensional space using a single-lens camera and measuring the displacement of a target object in the planar image in the XY directions. This technology is suitable for objects that need to be reliably returned to their test position. The device includes an image capturing module of a single-lens camera, a memory, and a processor. The processor performs steps such as image perspective transformation correction, structural similarity index calculation, multi-directional displacement detection, and result determination. The original image is transformed through perspective to generate a frontal image. The structural similarity index between the test image of the object and a standard image sample of the object is calculated. When the index value is lower than a preset threshold, X and Y direction scans are performed to determine the displacement and direction of the target object. This invention can be applied to quality control in the fields of panel manufacturing, semiconductor manufacturing, and precision machinery manufacturing, effectively improving inspection efficiency and reducing human inspection errors.

Claims

1. A detection device for detecting the displacement value of a target object in three-dimensional space in the XY direction using a single-lens camera, mainly comprising: an image acquisition module, which uses a single-lens camera to acquire an image of the position of the object to be tested; a memory for storing the image, detection area coordinates, and detection parameters; and a processor connected to the image acquisition module to acquire the image of the object to be tested, wherein the processor performs: (a) performing image perspective conversion correction processing on the original image of the object to be tested captured by the single-lens camera to generate a frontal image; (b) performing displacement detection calculation on the corrected frontal image and a standard image sample of the object to be tested, and obtaining a structural similarity index value; (c) when the aforementioned structural similarity index value is lower than a first preset threshold, performing a displacement scan to determine the displacement direction and displacement amount; and (d) comparing the aforementioned displacement amount with a preset allowable range to determine the displacement state and generate a detection result, and then storing the displacement value obtained from the detection result in the memory.

2. The detection device as claimed in claim 1, further comprising a user interface connected to the processor and the memory, primarily providing functions for displaying detection results and images, and setting detection parameters.

3. The detection device as described in claim 2, further comprising an output unit connected to the user interface for providing the function of exporting images and detection results.

4. The detection device as claimed in claim 3, wherein the processor further performs a statistical analysis process, which calculates the average displacement and the maximum displacement from the displacement data of multiple accumulated detection results, and stores the results in the memory for long-term trend analysis.

5. A method for detecting the displacement value of a target object in three-dimensional space in the XY direction using a single-lens camera, comprising the following steps: S1) Image perspective conversion correction – the original image captured by the single-lens camera is subjected to image perspective conversion correction processing to generate a frontal image; S2) Structural similarity index calculation – the structural similarity index value between the corrected frontal image and the standard image sample of the object to be tested is calculated; S3) Multi-directional displacement detection – when the calculated structural similarity index value is lower than a first preset threshold, displacement scanning is performed to determine the displacement direction and displacement amount; S4) Result determination – the displacement state is determined by comparing the displacement amount with a preset allowable range, and a detection result is generated.

6. The detection method as described in claim 5, wherein in step S1, the original image is converted into a frontal view by a preset four-point perspective transformation coordinate.

7. The detection method as described in claim 6, wherein the displacement detection in step S3 further includes a first stage detection and a second stage detection, wherein the first stage detection is to perform preliminary screening by comparing the structural similarity index value with the first preset threshold, and the second stage detection is to perform a stepwise scan in the X and Y directions on the region below the first preset threshold to determine the optimal matching position and displacement.

8. The detection method as described in claim 5, wherein the standard image sample of the test object is obtained by first correcting the test object through image perspective transformation to obtain a frontal image, then selecting the test object detection position in the corrected frontal image, and storing this position as the standard image sample of the test object; allowing the user to select multiple detection positions for the test object and store the coordinates of these detection positions to achieve simultaneous detection.

9. The detection method as described in claim 5, which includes a statistical analysis processing step, wherein the displacement data of multiple accumulated detection results are used to calculate the average displacement and the maximum displacement and the results are stored for long-term trend analysis.

Images