# US 12,603,968 B2 · Projection point extraction method > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,603,968 B2 - Document ID: US12603968B2 - Status: Granted - Kind: B2 - Application number: 18/631,096 - Filed: 2024-04-10 - Published: 2025-07-10 - Granted: 2026-04-14 - Earliest claimed: 2024-01-09 - Assignees: Primax Electronics Ltd. - Applicants: Primax Electronics Ltd. - Inventors: Hsiu-Wen Wang; Chih-Wen Lin - Web page: https://patentgrep.com/patent/US12603968B2 ## CPC - H04N 5/74: Projection arrangements for image reproduction, e.g. using eidophor - G06T 5/30: Erosion or dilatation, e.g. thinning - G06T 5/70: Denoising; Smoothing - G06T 7/11: Region-based segmentation - G06T 7/155: involving morphological operators - G06V 10/751: {Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching} - G06T 7/10: Segmentation; Edge detection (motion-based segmentation G06T7/215) - G06T 2207/20036: Morphological image processing - H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract A projection point extraction method for a detection system is provided. The detection system includes a projection module, a projection plane and a receiver. Firstly, an original image with plural projection points is projected. Then, a kernel is used to perform an erosion process on the original image, so that an erosion image is generated. Then, an intensity transformation process is performed on the erosion image according to an intensity threshold, so that a transformation image is generated. Then, a reconstruction process is performed on the transformation image according to the original image, so that a reconstruction image is generated. Then, a regional maxima process is performed on the reconstruction image, so that a mask map is generated. Afterwards, an extraction process is performed on the original image according to the mask map, so that a projection point extraction result is generated. ## Claims ### Claim 1 A projection point extraction method for a detection system, the detection system comprising a projector, a projection plane, a receiver, and a processor, the projection point extraction method comprising steps of: projecting an original image with plural projection points onto the projection plane by the projector, wherein after the original image is received by the receiver, the original image is transmitted to the processor; performing an erosion process on the original image by the processor using a kernel executing an operation program code to generate an erosion image; performing an intensity transformation process on the erosion image according to an intensity threshold by the processor executing the operation program code to generate a transformation image; performing a reconstruction process on the transformation image according to the original image by the processor executing the operation program code to generate a reconstruction image; performing a regional maxima process on the reconstruction image by the processor executing the operation program code to generate a mask map; and performing an extraction process on the original image according to the mask map by the processor executing the operation program code to generate a projection point extraction result. ### Claim 2 (depends on 1) The projection point extraction method according to claim 1, wherein the kernel is a matrix with a specified size, and corresponding elements of the matrix have identical or different contents. ### Claim 3 (depends on 1) The projection point extraction method according to claim 1, wherein the erosion process is a morphological processing process, wherein when the erosion process is performed, the kernel is used to scan the original image and suppress a surrounding background of the projection point in the scanned original image, so that a local maximum value of the corresponding projection point is obtained. ### Claim 4 (depends on 1) The projection point extraction method according to claim 1, wherein the erosion process comprises steps of: placing the kernel at any location of the original image, and allowing a center of the kernel to be aligned with one pixel of the original image; and using the kernel to scan the original image, eroding pixels of the original image at locations corresponding to elements of the kernel, and combining each element in the kernel and the corresponding pixel as a new pixel according to an erosion condition. ### Claim 5 (depends on 1) The projection point extraction method according to claim 1, wherein the intensity transformation process is a morphological processing process, wherein when the intensity transformation process is performed, an intensity of each projection point in the erosion image is suppressed according to the intensity threshold, and noise in the erosion image is filtered out. ### Claim 6 (depends on 1) The projection point extraction method according to claim 1, wherein the intensity transformation process comprises steps of: decreasing pixels in the erosion image that are greater than or equal to the intensity threshold; and decreasing or filtering out pixels in the erosion image that are less than the intensity threshold. ### Claim 7 (depends on 1) The projection point extraction method according to claim 1, wherein the reconstruction process is a morphological processing process, wherein when the reconstruction process is performed, each projection point in the transformation image is subjected to a dilation operation, so that characteristics of a corresponding region of the original image are restored. ### Claim 8 (depends on 1) The projection point extraction method according to claim 1, wherein the reconstruction process comprises steps of: setting an iterative image, wherein the iterative image has same dimensions as the original image, and all pixels of the iterative image are 0; and performing a dilation operation on the iterative image and the transformation image to obtain an operation result, and intersecting the operation result with the original image, so that the reconstruction image is generated. ### Claim 9 (depends on 1) The projection point extraction method according to claim 1, wherein the reconstruction image comprises plural preset regions, and the regional maxima process is a morphological processing process, wherein when the regional maxima process is performed, respective maximum values in the plural preset regions are determined, and noise in the reconstruction image is filtered out. ### Claim 10 (depends on 1) The projection point extraction method according to claim 1, wherein the reconstruction image comprises plural preset regions, and the regional maxima process comprises steps of: scanning all pixels of the reconstruction image; comparing a pixel value of each pixel in each preset region with pixel values of other surrounding pixels, wherein if the pixel value of the pixel to be compared is greater than or equal to all of the other surrounding pixels, the pixel value of the pixel to be compared is defined as a maximum pixel value in the corresponding preset region; and marking a location of the maximum pixel value in each of the plural preset regions. ### Claim 11 (depends on 1) The projection point extraction method according to claim 1, wherein the extraction process is an image processing process, wherein when the extraction process is performed, the mask map is used to perform a multiplication operation on the original image, so that the projection point extraction result is obtained. ### Claim 12 (depends on 1) The projection point extraction method according to claim 1, wherein the detection system further comprises a flash memory storing the operation program code, the flash memory is electrically connected with the processor, and the projector, the receiver, and the processor are integrated as a projection device. ### Claim 13 (depends on 1) The projection point extraction method according to claim 1, wherein the receiver is a camera. ### Claim 14 (depends on 8) The projection point extraction method according to claim 8, wherein the reconstruction process further comprises a step of continuously iterating the generated reconstruction image until the newly generated reconstruction image is no longer changed or a predetermined iterative value reaches. ### Claim 15 (depends on 10) The projection point extraction method according to claim 10, wherein the other surrounding pixels for the pixel to be compared include eight neighboring pixels in an upper side, a lower side, a left side, a right side and four diagonal directions of the pixel to be compared. ### Claim 16 (depends on 10) The projection point extraction method according to claim 10, wherein in the regional maxima process, the pixel value of the pixel defined as the maximum pixel value in the corresponding preset region is set as 1, and the pixel value of the pixel that is not defined as the maximum pixel value in the corresponding preset region is set as 0.