The invention, which belongs to the technical field of measurement, particularly relates to a detection method for a trace amount of dust based on microscopic image processing. The method comprises the following steps: step 1, collecting an image of eachdust particle on a certain surface by using a microscopic imaging system; step 2, performing gray processing on the dust particle images; step 3,performing illumination correction ongrayscale images by using a block background extraction method; step 4, carrying out median filtering processing on the images after illumination correction; step5, drawing gray histograms for the grayscale images; step 6, carrying out image edge detection on the images by adopting a Prewitt operator; step 7, performing binary image processing such as expansion and corrosion on the images; step 8, performing feature extraction on the images, and carrying out statistics on the number of pixel points occupied by each dust particle so as to obtain a pixel area of each dust particle; and step 9, deducing the volume of the dust particle according to the pixel area of the dust particle and acquiring the mass of each dust particle based on the volume and density to obtain the total mass of the dust particles.

Detection method for trace amount of dust based on microscopic image processing

The invention discloses an image feature descriptor extraction method and device, equipment and a storage medium, and relates to the field of image recognition. The method comprises the steps of obtaining a to-be-processed image; expanding the edge of the to-be-processed image outwards by adopting the pure color pixel points to obtain a filling image; extracting an image feature descriptor of thefilling image; enabling the box type filter to extract the features of the edge pixel points of the image to be processed, extracting the features in the image to be processed fully, and effectivelyavoid the edge effect. For a small-size image, enough features can be accurately extracted from the processed filling image through an SIFT/SURF algorithm.

Image feature descriptor extraction method and device, equipment and storage medium

The invention relates to the technical field of image processing, and provides a method and a device for image coding and logical operation The method comprises the following steps: firstly, extending multiple binary images into multiple square binary images respectively; secondly, according to a node code q and a pixel code r of a node in each square binary image, generating a truncation quadtree code of each square binary image; thirdly, performing logical operation on the truncation quadtree code to obtain a result coded sequence; and finally, performing decoding according to the result coded sequence to obtain an extended binary image of 2N*2N, and capturing a sub-image extended by the extended binary image to obtain a result binary image Therefore, a binary image code of which the space overhead is small enough can be obtained, and meanwhile, the speed is higher when the logical operation of the binary image is carried out

Method and device for image coding and logical operation

The invention discloses an adipocyte profile parameter image measuring method. After the pork adipocyte sample is prepared, the microscopic image is acquired, through the image processing algorithm, programs are written to automatically segment the adipocyte, and on this basis, the pork adipocyte parameter measurement can be automatically completed, the labor intensity is less, the work efficiency is high, the result is stable, the accuracy is high, and the error result caused by fatigue or subjective determination factors in the manual operation can be prevented.

Adipocyte profile parameter image measuring method

The invention provides a learning method and device for an electricity consumption pattern as well as an intelligent ammeter, and relates to the technical field of intelligent appliances. The method comprises the following steps: obtaining the electricity consumption corresponding to actual situation atmospheric temperature in a preset learning period; according to the electricity consumption and the preset temperature range of the actual situation atmospheric temperature, establishing an electricity consumption sample set of the preset temperature range; according to an inverse function of a normal distribution probability function and the number of samples in the electricity consumption sample set, calculating a first parameter set; and according to the electricity consumption sample set and the first parameter set, establishing an electricity consumption pattern learning model, so as to realize the learning of an electricity consumption pattern of the preset temperature range. Through the adoption of the learning method disclosed by the invention, the electricity consumption habits of a personal user in different preset temperature range are analyzed, so that according to the electricity consumption habits of the personal user, privately-customized precision service can be provided.

Song Pengcheng

Papers

Method and device for image coding and logical operation

Adipocyte profile parameter image measuring method

Learning method and device for electricity consumption pattern, as well as intelligent ammeter