| CPC G06V 10/95 (2022.01) [G06Q 10/20 (2013.01); G06T 5/73 (2024.01); G06V 10/44 (2022.01); G06V 20/46 (2022.01); H04N 7/183 (2013.01); G06T 2207/20021 (2013.01); G06T 2207/20192 (2013.01); H04W 4/30 (2018.02); H04W 4/80 (2018.02)] | 5 Claims |

|
1. An intelligent mobile patrol method, comprising:
via a processor, sending a patrol task to an intelligent mobile device bound with Near Field Communication (NFC);
via the intelligent mobile device, identifying the patrol task to collect patrol data including on-site photos, recordings, and videos, wherein the patrol data is accumulated as on-site visual data;
via the processor, instructing the intelligent mobile device to upload the on-site visual data to a memory for intelligent processing and analysis operation, so as to query historical patrol records in real time and analyze a maintenance rate, a missing inspection rate and an equipment availability rate;
via the processor, instructing the intelligent mobile device to upload point patrol records to the memory for storage in real time by using WiFi, and automatically import equipment exceptions into a missing library in the memory; and
via the processor, consolidating and transmitting an obtained patrol record data to the intelligent mobile device, so as to push a real-time patrol task and receive patrol results;
wherein the intelligent processing and analysis operation comprises:
via the processor, performing image preprocessing operation on screen information of the on-site visual data to obtain processed visual data;
via the processor, performing features extracting operation on the processed visual data to obtain corresponding feature data in the screen;
via the processor, identifying the feature data and comparing the feature data with data in database; and
via the processor, according to comparison and screening results, extracting historical data of corresponding database for data accounting, and updating the patrol data to form a new patrol task;
wherein: the image preprocessing operation comprises:
transforming each pixel gray scale of the visual data to expand the gray scale range of a visual data image;
improving the visual data clarity to balance the spatial pixel value uniformity of the visual data image; and
performing fuzzy enhancement at an edge of pixel, so as to strengthen boundary distinction;
the features extracting operation comprises:
via the processor, adopting gamma correction method to standardize the color space of visual data image;
via the processor, calculating gradient of each pixel of the visual data image to capture the contour information and weaken the interference of light;
via the processor, dividing the visual data image into small cells to count the gradient histogram of each small cell;
via the processor, forming every several small cells into a block to obtain a feature block of a block; and
via the processor, concatenating the feature block to obtain feature data of the visual data image for classification.
|