【摘要】 提出了一种对彩色水果图像进行果实轮廓提取的算法。首先将图像转换为OHTA颜色空间中的灰度图像,用阈值法实现了图像分割;然后用BLOB算法滤去图像中的噪声;最后用插值的方法得到平滑的轮廓曲线。利用开发的智能瓜果精选分级试验样机,按照大小对温室西红柿进行分类,对算法的精度和实时性进行检验,结果表明算法的精度可达98%,分选速度达到3个/s,达到了实用化的目标。更多还原

【Abstract】 An algorithm to realize fruit contour detection in color image is presented in this paper. At first, a color image is transformed into a grey-level image in OHTA color space, and segmented by constant threshold value; then the BLOB algorithm is utilized to erase noises in the image; a spline interpolation based algorithm is adopted to get smooth fruit contour at last. To test the speed and accuracy of algorithm, 50 green-house tomatoes were classified according to size in the exploited intelligent fruit sorting experimental prototype. The experimental result shows that the accuracy of the algorithm is higher than 98%, its calculation speed is 3 samples per second, which satisfies practical requirements.更多还原