Image Processing Apparatus for Object Separation Using 2D and Distance Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems using 2D cameras for object recognition can lead to erroneous recognition due to surface patterns, while 3D cameras require high resolution to separate closely arranged objects, which is costly and inefficient.
Innovation Solution
An image processing apparatus that captures and processes both 2D and distance images, extracts pixels with low brightness variation across different exposure conditions, and sets corresponding pixels in the distance image as non-imaging pixels to improve object separation and recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a two-dimensional camera is used for object recognition, then the system cost is reduced, but erroneous recognition occurs due to surface patterns affecting contour extraction
Solution Approach 1:
The patent combines 2D image data from a standard camera with distance image data from a time-of-flight camera to create composite image data. This merging allows the system to use the contour information from the 2D image while correcting for surface pattern interference using the depth information from the distance image, thereby maintaining low system cost while improving recognition accuracy.
Solution Approach 2:
The patent introduces distance image data as an intermediary to mediate between the 2D image and the final recognition result. The distance information acts as a mediator that helps distinguish actual object contours from surface patterns by providing depth context, allowing accurate recognition without requiring expensive high-resolution 3D cameras.
2Reliability
If a three-dimensional camera is used to avoid surface pattern influence, then recognition accuracy is improved, but high resolution is required to separate closely arranged objects which increases cost
Solution Approach 1:
The patent merges low-resolution distance image data with 2D image data to achieve the separation of closely arranged objects. By combining the depth information (even at low resolution) with the detailed 2D image information, the system can identify narrow gaps between objects without requiring a high-resolution 3D camera, thus reducing system cost while maintaining recognition accuracy.
Solution Approach 2:
The patent creates a composite image data structure that combines 2D image data and distance image data. This composite approach allows the system to leverage the strengths of both data types: the detailed visual information from the 2D image and the depth separation information from the distance image, achieving accurate object separation without the need for expensive high-resolution 3D imaging.
3Ease of manufacture
If low resolution distance image is used, then system cost is reduced, but closely arranged objects cannot be separated by narrow gaps
Solution Approach 1:
The patent combines low-resolution distance image data with high-resolution 2D image data to achieve precise object separation. The distance image provides coarse depth information that helps identify potential gaps between objects, while the 2D image provides detailed visual information to confirm object boundaries, together enabling separation of closely arranged objects without requiring high-resolution 3D imaging.
Solution Approach 2:
The patent adds the depth dimension from distance image data to the 2D image data, creating three-dimensional composite image data. This dimensional enhancement allows the system to separate closely arranged objects by utilizing depth information even when the distance image resolution is low, as the additional dimensional information provides the necessary discrimination between adjacent objects.
Data Source
AI summary
An image processing apparatus includes a two-dimensional image storage unit configured to store a plurality of two-dimensional image data captured by photographing an identical imaging target object under different exposure conditions; a distance image storage unit configured to store distance image data including a pixel array of a known relationship to a pixel array of the two-dimensional image data; a pixel extraction unit configured to extract, among pixels in each of the two-dimensional image data, a first pixel at which a difference in brightness between identical pixels is less than a predetermined value; and a distance image adjusting unit configured to specify a second pixel of the distance image data at a position corresponding to the first pixel in the pixel array, and to set the second pixel as a non-imaging pixel in the distance image data.