Artificial intelligence corn quality detection robot and quality detection method

Abstract

The invention provides an artificial intelligence corn quality detection robot and a quality detection method. The artificial intelligence corn quality detection robot comprises an artificial intelligence corn quality detection robot main body and a system server, the artificial intelligence corn quality detection robot main body comprises a box body, the box body is provided with a material inletand a material outlet and is internally provided with a double-way material feeding mechanism, a conveying mechanism, an up and down double-faced machine vision mechanism and a double-way sorting mechanism, the material inlet communicates with the double-way material feeding mechanism, the double-way material feeding mechanism communicates with the conveying mechanism, the conveying mechanism passes through the up and down double-faced machine vision mechanism and communicates with the material outlet, the double-way sorting mechanism is located at the rear of the up and down double-faced machine vision mechanism, the up and down double-faced machine vision mechanism and the double-way sorting mechanism are connected with the system server, and the double-way material feeding mechanism adopts the method of a sensor and intelligent control. The invention further comprises the corn quality detection method. By means of the artificial intelligence corn quality detection robot and the quality detection method, the problem that artificial sensory detection is affected by human factors and external environment to cause detection index drift can be effectively solved, and rapid detectionis realized.

Description

technical field [0001] The invention relates to corn quality physical testing equipment and methods, in particular to an artificial intelligence corn quality inspection robot and a quality inspection method. Background technique [0002] At present, there are no modern means and equipment for corn detection, and corn detection is completed by artificial sensory identification, and there are many uncertain factors in artificial sensory identification. For example, the same sample is identified multiple times by the same inspector, and the data of each identification will be different. For the same sample, if different inspectors use the same standard, the identification data will be different; therefore, it is difficult to distinguish objectively, fairly and accurately, and it has become a fact that "there are standards but no technical means to implement the standards". [0003] The existing artificial sensory recognition can complete the detection of imperfect corn kernels,...

AI Technical Summary

Problems solved by technology

[0002] At present, there are no modern means and equipment for corn detection, and corn detection is completed by artificial sensory identification, and there are many uncertain factors in artificial sensory identification. For example, the same sample is identified multiple times by the same inspector, and the data of each identification will be different. For the same sample, different inspectors use the same standard, and the identification data will be different; therefore, it is difficult to distinguish objectively, fairly and accurately, and it has become a fact that "there are standards but no technical means to implement the standards"

[0003] The existing artificial sensory recognition can complete the detection of imperfect corn kernels, and has the following disadvantages: 1) It is easily disturbed by environmental and human factors; 2) There is no unified measurement tool; 3) Inefficiency

[0005] (1) The feeding method of the screw mechanism is usually only suitable for standard parts. When the material is a non-standard part of grain, it is prone to blockage, leakage, Incomplete discharge and other conditions, thus making the transported samples incomplete and seriously affecting the efficiency

[0006] (2) Optical glass disc feeding method. When the disc rotates to a certain speed, the material is easy to slip off. At the same time, due to the end jump generated when the disc rotates, the samples sent into the visual field lack stability, which further makes the acquisition The measurement space of the material is distorted, and the original information of the detected object is seriously deviated

[0007] (3) The relative visual mode of the upper and lower sides causes the upper and lower visual information to interfere with each other, which further causes the original information of the detected object to be distorted and produces greater deviations

[0008] (4) The mechanical comparison between the original image of the detected object and the image library of the standard sample cannot adapt to the non-standard feature recognition of grain particles, and the classification of imperfect grain particles does not follow the linear relationship, and the mechanical comparison is used for rigid linear relationship discrimination In actual work, the error is very large

[0009] (5) When the detected particles arrive at the sorting area from the identified area, the physical position is often due to sliding and disc end jumping, so that the sorting accuracy cannot be guaranteed

Images