Light Object Image Capture Without Edge Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image capture techniques struggle to capture images of light-colored objects on light-colored backgrounds, as they require identifying the object's edges, which is not feasible in such scenarios.
Innovation Solution
A system and method that allow capturing and processing images of objects without identifying their edges, by converting the image to a lower quality greyscale, filtering out noise, and defining a rectangle around identified edges. The image is then checked for acceptable aspect ratio, margin, and area occupancy before being uploaded to a server for further processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional image capture techniques are used that identify object edges, then image capture of objects with distinct boundaries is achieved, but image capture of light-colored objects on light-colored backgrounds fails
Solution Approach 1:
The patent converts the captured color image to a lower quality greyscale image, fundamentally changing the parameter of image representation from color to grayscale. This transformation enables the system to detect edges and boundaries even when the object and background have similar light colors that are difficult to distinguish in the original color space.
Solution Approach 2:
The patent replaces the conventional edge identification mechanism with an alternative approach that captures images without requiring precise edge detection. Instead of relying on edge-based algorithms that fail for light-on-light scenarios, the system uses a different technical pathway involving greyscale conversion and noise filtering to achieve successful image capture and upload.
2Measurement precision
If high quality color images are captured and uploaded, then image quality is maintained, but processing time and data usage increase
Solution Approach 1:
The patent applies partial action by converting the full quality color image to a lower quality greyscale version for edge detection purposes. This partial conversion allows the system to perform necessary processing with reduced data requirements, while still capturing sufficient information to identify whether the image meets upload criteria.
Solution Approach 2:
The patent segments the image processing task into distinct stages: first converting to greyscale for edge detection, then evaluating whether the captured image meets quality thresholds for upload. This segmentation allows the system to use lower quality processing for preliminary evaluation and reserve full quality uploads only when necessary, reducing overall data transmission and processing time.
Data Source
AI summary
A system and method allows a light colored image of an object such as a check to be detected and captured on a light colored background for uploading to a server for processing. Detection involves detecting edges of objects on the image, drawing a rectangle around the detected edges, testing for an aspect ratio of the rectangle within an approved range, testing for the rectangle being outside of a margin of the image and being a certain percentage of the image, and testing for blur within a tolerable range.