A functional diagram of a robotic complex mock-up for rejecting ball-shaped objects transported on a roller conveyor is presented. The software of the complex is designed to detect an object on the conveyor and determine its coordinates. To detect objects of control, their images obtained in the visible range of the optical spectrum and Viola–Jones object detection algorithms are employed. The developed software is based on a trained cascade classifier, the optimal settings of which are determined. To detect surface defects of objects, their spectrograms obtained in the range of 400… 1000 nm are used. The presented results can be applied to the process automation of sorting fruits and vegetables and other ball-shaped objects.

No abstract

The article is devoted to the development of algorithms for detecting defective apples transported on a roller conveyor using a vision system. In developing the algorithms, the possibility of classifying various regions of interest (intact and damaged by rot, scab, codling moth, as well as the conveyor) by the principal component method was investigated. When choosing the optimal spectral region for cluster analysis, spectrograms obtained in various spectral ranges, including Vis-NIR (400–1000 nm), NIR (780–1000 nm), and Vis (400–780 nm) were used. The PCA method showed that for the successful classification of the conveyor area, intact, decayed and damaged by the codling moth, it is necessary to use spectrograms in the Vis-NIR range. To classify these ROIs, it was proposed to use a direct distribution neural network with two hidden layers of 128 and 64 layers, respectively, the “relu” activation function in the hidden layers and the “softmax” activation function in the output layer. The optimal network configuration was determined experimentally. This configuration showed a classification accuracy of 0.847 on a test sample of 6,000 apples. Since the samples of spectrograms of scab and stem regions do not differ, for their classification in parallel with the neural network, it was proposed to use the Haar cascade classifier trained on 2000 two-dimensional images of apples in the visible region containing scab and stem regions. The classification accuracy was at least 0.95. The developed algorithm is intended for use in the robotic sorting of apples.

The functional diagram of the mechatronic system for sorting and monitoring the quality of vegetables and fruit is given. The control objects were exposed to short-term thermal effects from sources of IR radiation during the movement on a roller table conveyor. The temperature fields of healthy control objects and damaged one by phyto-diseases or mechanical effects are different. This fact allows using a vision system based on a thermal imaging camera to detect subsurface defects for rejecting damaged control objects. Hyperspectral method of control was used for the surface defects detection. Defects detection algorithms in IR range images, as well as on the results of spectrogram processing, are given.

No abstract