Three-dimensional analysis of fruit and vegetables: Sensors and methods

Measuring physical traits like volume, size, shape, and uniformity is crucial for food identification, grading, and categorisation. However, conventional non-contact methods, which primarily rely on visual assessments and two-dimensional image processing, face significant challenges when accurately evaluating complex fruit characteristics. The increasing accessibility and continuous improvement of three-dimensional vision technologies offer substantial potential for developing automated systems that can visualise and reconstruct food morphology. By integrating 3D computer vision techniques with advanced image processing, we can achieve accurate morphological measurements. To date, researchers have explored diverse sensor technologies for image processing-based three-dimensional reconstruction, identifying optimal monitoring techniques for a wide range of fruits and vegetables. This paper critically reviews recent advances in food reconstruction. It expands on existing research by clarifying target development, parameters, sensor technologies, and the strengths and weaknesses inherent in three-dimensional reconstruction. This review emphasises the benefits of combining two-dimensional image processing with artificial intelligence algorithms in three-dimensional reconstruction, particularly in improving measurement accuracy, reducing costs, and simplifying system complexity. Ultimately, this paper aims to guide further research in food inspection and quality control.