Research on Superficial Area Measurement Method of Irregular Object Based on Image Processing

Yi Yu2016 · DOI

Abstract

Every field needs to face the problem of superficial area measurement of irregular objects. How to quickly, effectively, and accurately measure the superficial area of irregular objects has a positive effect on increasing economic benefits, attracting universal attention. With the development and application of computer digital image technology, excellent achievements have been obtained in different fields. For this, this paper puts forward a superficial measurement method for irregular objects based on image processing. It separates the target area and background area by collecting the gray-level image of the object and combining it with a threshold image segmentation method. It then uses boundary following to obtain information such as boundary points of the target area, tracking direction, and pixel count, allowing for accurate measurement of the superficial area. This method offers simple operation and high accuracy, giving it significant application and popularization value.

Introduction

In industrial sectors such as leather processing, textile manufacturing, and metal fabrication, calculating the surface area of irregular shapes is a persistent challenge. Traditional manual methods are often slow, prone to human error, and difficult to scale. Accurate area measurement is critical for "active cost accounting"—optimizing material usage to reduce waste and increase economic efficiency.

This paper presents a non-contact, image-based system designed to address these needs. By combining adaptive threshold segmentation with an 8-direction boundary tracking algorithm, the proposed method achieves a measurement precision of ±1%, offering a significant improvement over traditional techniques.

Methodology

Hardware Configuration

The system utilizes a custom hardware setup designed for stability and precision. A CCD camera is mounted on a flexible slide bar, allowing for adjustable distances between the lens and the object. To ensure consistent results, a dedicated lighting sub-system provides uniform illumination, minimizing shadows and background noise that could interfere with image segmentation.

The hardware system setup including CCD camera, lighting, and processing unit.

Software Processing Pipeline

The software component, developed in Visual C++ 6.0, follows a multi-stage pipeline to extract area data from raw images.

Flowchart of the image processing pipeline from data collection to area calculation.
  1. Preprocessing: Color images are converted to gray-scale. A median filter is then applied to remove salt-and-pepper noise while preserving critical edge information.
  2. Adaptive Thresholding: The system iteratively calculates an optimal threshold ($T$) by finding the mean gray levels of the target area ($\mu_1$) and the background ($\mu_2$). This ensures robustness against varying lighting conditions.
  3. Boundary Tracking: Using an 8-neighbor connectivity model (Wang Houda’s eight-direction code), the system traces the object's perimeter to identify boundary points and tracking directions.
  4. Area Calculation: The area is calculated using a formula that integrates internal and boundary pixels: $$Area = \left( \frac{\text{Total boundary pixels}}{2} + \text{Total pixels within boundary} \right) \times \text{pixel equivalent coefficient}$$ The "pixel equivalent coefficient" is a calibration factor determined by measuring an object of known dimensions, bridging the gap between digital pixels and physical units (cm²).

Results and Evaluation

The method was evaluated using four specific samples of irregular leather to test its performance on complex, organic shapes. The results demonstrated high precision and efficiency:

  • Leather 1: 1,388 boundary pixels; 121,465 internal pixels; Area = 1,476.287 cm².
  • Leather 3 (Largest): 2,324 boundary pixels; 299,678 internal pixels; Area = 3,635.641 cm².
  • Precision: The system achieved a measurement precision of ±1%.

The "Half-Boundary" rule in the area formula (dividing boundary pixels by two) proved essential for accuracy, as it accounts for pixels that straddle the edge between the object and the background, preventing overestimation.

Conclusion and Future Work

The proposed image processing method provides a simple, fast, and accurate solution for measuring irregular surface areas in industrial environments. While highly effective for 2D projections, the current system's primary limitation is its inability to account for 3D surface variations (e.g., wrinkles or curves). Future work could involve integrating depth-sensing technology to extend these measurements into three-dimensional space, further enhancing accuracy for non-planar materials.

Summary updated

Abstract

every field needs to face with problem on superficial area measurement of irregular object, how to quickly, effectively and accurately measure the superficial area of irregular object has appositive effect on increasing economic benefit, so it attracts universal attention. With the development and application of computer digital image technology, it has obtained excellent achievement in the different fields. For this, this paper puts forward one kind of superficial measurement method based of irregular object based on image processing, it separates target area and background area through collecting the gray level image being measured and combining with threshold image segmentation method. And then it uses boundary following to obtain information such as boundary point of target area as well as tracking direction and pixel count of boundary point, so it can well measure superficial area of irregular object. This method has advantage of simple operation and high accuracy, so it has higher application and popularization values.
BibTeX
@misc{yu2016research,
  author = {Yi Yu},
  title = {{Research on Superficial Area Measurement Method of Irregular Object Based on Image Processing}},
  year = {2016},
  doi = {10.2991/icence-16.2016.153},
}
Figure