Tobacco stem content detection method based on image method based on multi-feature fusion

Abstract

The invention discloses an image method based on multi-feature fusion to detect the tobacco stem content rate, which includes using the three texture features of the original image, the highlight density, the gray level entropy and the gray level, to construct the probability density image of the tobacco stem area, and then adopting bilateral The filter function filters the probability density image, and finally uses the mixed Gaussian model and morphological filtering method to extract the accurate tobacco stem area from the tobacco stem probability density image; the tobacco leaf image obtained after image processing uses the fitting algorithm to calculate the stem content rate . The invention suppresses noise points in the tobacco stem X-ray image and the interference of pseudo tobacco stems formed by overlapping edges, significantly improves the recognition accuracy of the tobacco stem region, and improves the accuracy of calculating the stem content rate.

Description

technical field [0001] The invention relates to an image-based tobacco stem content detection method based on multi-feature fusion, and belongs to the technical field of tobacco stem content detection technology. Background technique [0002] In the process of threshing and redrying tobacco leaves, it is necessary to use air separation to separate the threshed tobacco sheets and tobacco stems, so as to carry out accurate formulation of tobacco leaves in the later stage and improve the quality of cigarettes. In the blended tobacco leaves after threshing, there is little difference in the suspension velocity of stem-containing tobacco sheets and free pure leaves. During the air separation process, the stem-containing tobacco sheets will be regarded as pure leaves and difficult to separate, so that the tobacco leaf formula The addition of stem-containing tobacco slices directly affects the quality of cigarettes. The stalk content rate of tobacco leaves after threshing is one o...

AI Technical Summary

Problems solved by technology

[0003] According to the above national standards, each threshing and re-baking enterprise must test the content of tobacco stems. Most re-baking enterprises have to test the content of tobacco leaves every half an hour. There are two main measurement methods: one is, Use a leaf thresher to smash all the tobacco leaves to be tested, and then measure the stem content rate. This measurement method will cause damage to the tobacco leaves. After the measurement, the tobacco leaves will not be able to continue to be used. Every day, every threshing line will be wasted. a lot of tobacco leaves

Second, at present, most tobacco factories and rebaking factories measure the stem content of tobacco leaves by weighing the tobacco stems and leaves after sampling and destemming. This detection method causes great damage to the tobacco leaves, and the detection takes a long time and Unable to realize online real-time detection

Both of these detection methods have obvious disadvantages

Images