Cane stalk recognition method

A recognition method, sugarcane stem technology, applied in the field of recognition, can solve problems such as the sugarcane stem node pattern recognition method that has not yet been seen, and achieve the effects of reducing the rate of bud damage, accurate judgment, and improving labor productivity

CN102663434AInactive Publication Date: 2012-09-12GUANGXI UNIV
0 Cites 5 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2012-09-12
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a cane stalk recognition method which is based on computer vision. The method includes that after k adjacent given samples [xi, i=1,2,...,k] of a cane image sample x containing stalk joints and stalk sections to be sorted are obtained, membership degree value of the sample x on each category is calculated; then total sum of the membership degree of all samples on each category is calculated to be n, membership degree of a sample xi to be sorted on j category according to mu j(x) is calculated and compared, a category where the maximum membership degree belongs is selected so as to judge whether a sample is a stalk joint or a stalk section. The method provides reference for cane planting machinery which is not provided with a bud-damage resisting device yet at home and abroad, fills an application gap of image processing technology in cane seed stalk cutting at home and abroad, uses the computer image processing technology to achieve fast and accurate judgment of a cutter for cutting cane portions, reduces bud damaging rate of cane seeds and stalks and cane production cost, saves cane seeds, and improves labor productive power.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to a recognition method, in particular to a sugarcane stem node recognition method. Background technique

[0002] my country is the world's largest producer of sugarcane, and as the third largest producer of sugar in the world, the development of sugarcane planting directly affects the livelihood of tens of millions of sugarcane farmers and the development of the sugar industry. Most of the world's sugarcane producing areas have realized the mechanization of sugarcane planting to a certain extent, but there are deficiencies. Although foreign planters have good performance and perfect functions, they are not yet equipped with a professional anti-injury bud cutting device, and the mechanism is too complicated and the price is too expensive, so it is difficult to promote in domestic sugarcane production areas. However, domestic planters are more difficult to realize the purpose of automatically preventing damage to buds in the proce...

Examples

Embodiment

[0027] Fuzzy k-nearest neighbor method was used to identify the collected images of sugarcane stem nodes, and 50 sugarcane image samples with 1-3 stem nodes were selected. According to the combination of three characteristic parameters of sugarcane edge curve smoothness, stem node surface color and stem node leaf scar shape, it is used as the feature vector for subsequent pattern recognition. In order to identify the stem nodes in the sample, the fuzzy k-nearest neighbor method is used here to take them as input vectors. In order to prevent excessive exaggeration or reduction of the effect of a certain feature, the feature value of the input sample needs to be normalized. The algorithm normalizes the eigenvalues ​​of the samples to [0, 1].

[0028] The mean vector of stem nodes and internodes in the sugarcane sample is:

[0029] m d 1 = 1 10 Σ ...