Method and device for counting number of target objects in region

Abstract

The invention provides a method and device for counting the number of target objects in a region. The method comprises the steps of collecting images in the region; extracting multiple images in the images to serve as key frame images; screening the multiple key frame images, calibrating target objects in the key frame images by adopting a calibration frame; operating according to the calibrationfor the target objects in the key frame images to obtain a preliminary detection result of the number of the target objects; performing first tracking inspection on the preliminary result; performingsecond tracking inspection on the preliminary result; updating the number of the target objects according to the first tracking inspection result and the second tracking inspection result so as to obtain a final detection result of the number of the target objects. According to the method provided by the invention for counting the number of target objects in a region, data tracking is performed onthe data twice, and the number of the target objects is updated through the mutual cooperation of the two tracking inspection results, thereby accurately obtaining the number of the targets, and greatly improving the result sudden change and deterioration caused by the jitter and overlap of the target objects.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method for counting the number of objects in an area and a device for counting the number of objects in the area. Background technique [0002] The description of the background technology of the present invention belongs to the relevant technology related to the present invention, and is only used to illustrate and facilitate the understanding of the content of the present invention. prior art on the filing date. [0003] At present, most breeding enterprises rely on manual inventory and calculation, not only the working environment is harsh, but also the counting of real-time moving animals will inevitably produce errors. At the same time, the low labor efficiency further reduces the timeliness of data and further reduces the value of data. Among the emerging solutions, a large number of neural network methods are limited to the training and modification of detectors, and fo...

AI Technical Summary

Problems solved by technology

[0003] At present, most breeding enterprises rely on manual inventory and calculation, not only the working environment is harsh, but also the counting of real-time moving animals will inevitably produce errors

At the same time, the low labor efficiency further reduces the timeliness of data and further reduces the value of data

Among the emerging solutions, a large number of neural network methods are limited to the training and modification of detectors, and for a large-scale and large-scale farming scene, a single camera cannot cover all areas, and a single target detection algorithm counts The time series and the spatial mobility of livestock will also be ignored in the process, and the results will be greatly deviated

In the early research work, the matching problem of detection is the most important core problem of tracking problem. Therefore, based on the traditional Hungarian algorithm, the effect of Kalman filter algorithm combined with cascade matching will be affected by sudden changes in detector performance and overlapping objects. leading to strong degradation in tracking results

Images