Optical Pen Marker Segmentation for Low-Power 2D Code Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting multiple two-dimensional codes in an image require substantial processing power and time, are inefficient in cost and speed, and fail to accurately locate and decode spaced-apart markers in scenarios with low processing resources.
Innovation Solution
A method involving image segmentation into sub-areas using statistical data analysis, including rolling mean and variance, to identify and isolate individual two-dimensional markers, allowing efficient processing without extensive resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple separate applications are used for different marker detection functions (camera-based detection, RFID detection, barcode detection), then each function can be optimized independently, but the overall system complexity increases and requires multiple devices
Solution Approach 1:
The patent combines camera-based detection, RFID detection, and barcode detection into a single integrated application. The server receives data from multiple sensors (camera, RFID reader, barcode scanner) and processes all marker types uniformly, eliminating the need for separate applications while maintaining detection accuracy for each marker type.
Solution Approach 2:
The server is designed with universal processing capability to handle multiple marker types (spatially-distributed 2D markers, RFID tags, barcodes) through a single interface. The same server infrastructure processes all marker data regardless of source, enabling one application to perform functions previously requiring multiple specialized applications.
2Measurement precision
If spatially-distributed 2D markers are used instead of traditional center-point markers, then marker detection accuracy and robustness to occlusion improve, but the complexity of marker processing and coordinate transformation increases
Solution Approach 1:
The patent creates a virtual model (copy) of the physical workspace based on detected spatially-distributed markers. The server processes the coordinates of multiple marker points to generate a virtual representation of the real space, including virtual objects and their spatial relationships. This virtual model simplifies subsequent processing by pre-establishing coordinate transformations and spatial mappings.
Solution Approach 2:
The patent replaces complex manual coordinate transformation calculations with an automated server-based processing system. Instead of requiring real-time mathematical transformations during device operation, the server pre-processes marker coordinates and establishes transformation relationships, substituting mechanical calculation complexity with centralized computational processing.
3Productivity
If real-time collaborative editing is implemented across multiple terminals, then user productivity and collaboration efficiency improve, but network bandwidth consumption and server processing load increase
Solution Approach 1:
The server implements periodic synchronization of workspace data to connected terminals rather than continuous real-time streaming. Data is updated and pushed to terminals at regular intervals or when changes occur, reducing network bandwidth consumption while maintaining collaborative editing functionality. This periodic action balances productivity requirements with energy efficiency.
Data Source
Figure 1
Figure 2
Figure 2A~2B
AI summary
There is described a method of detecting and processing spaced-apart two-dimensional markers (C) that are distributed over a surface (S) to be imaged by means of an image acquisition sensor (10). The method comprises the following steps, namely, (a) acquiring an image (I) of at least a portion of the surface (S) by means of the image acquisition sensor (10), which image (I) is expected to include multiple ones of said two-dimensional markers (C), (b) segmenting the image (I) acquired at step (a) into multiple sub-areas each containing a single two-dimensional marker (C), and (c) subjecting one or more of the sub-areas that contain a two-dimensional marker (C) to further processing. Segmenting the image (I) at step (b) includes (b1) sampling the image (I) at a coarse sampling resolution lower than a resolution of the image (I) to retrieve statistical data indicative of a distribution of the two-dimensional markers (C) in the image (I) acquired at step (a). Segmenting (b) also includes (b2) processing the statistical data retrieved at step (b1) to identify expected intermediate areas (AB) separating the two-dimensional markers (C) and expected marker areas (AC) containing the two-dimensional markers (C). Segmenting (b) further includes (b3) generating a segmentation mask (MSEG) based on the identification of the expected intermediate areas (AB) and marker areas (AC) at step (b2). Lastly, segmenting (b) includes (b4) segmenting the image (I) acquired at step (a) into the multiple sub-areas based on the segmentation mask (MSEG) generated at step (b3).