Uncleanness detecting device

Abstract

The inventive uncleanness detecting method uses a subject region extracting unit to automatically extract a subject to be compared object in the image from an image picking-up device that is fixed at a definite point and has no movable system and to further extract an image region from the inside of the subject. A plurality of the latest extracted image regions extracted the image region from the inside of the subject is held in an extracted region image memory unit. Furthermore, a comparison of the extracted region images held by an uncleanness calculating unit is carried out. In the uncleanness calculating unit, each of the extracted region images is compared and, when the uncleanness calculating unit shows a high possibility situation of attached uncleanness, the degree of cleanness for its corresponding pixels stored in an uncleanness memory unit is increased, and vise versa. Every time an image is picked up, a subject region extraction from the image and calculation of the uncleanness thereof are carried out, so that information stored at uncleanness memory unit is updated. An uncleanness judging unit outputs a judgment result as to whether uncleanness is attached or not, or how the degree thereof is, in light of uncleanness information stored the uncleanness memory unit.

Description

technical field [0001] The present invention relates to a technique for detecting dirt or foreign matter adhering to the surface of an image sensor that is exposed outside a device that captures an image and is used in a fixed-point surveillance camera or an inspection device. Background technique [0002] In the use of a surveillance camera or an inspection device, it is very important to capture an image as vivid as possible within the performance range of an image sensor. However, dirt or foreign matter adhering to the surface of the sensor exposed to the outside of the device will hinder the capture of clear images, and processing images captured under such conditions may cause device malfunction or the like. [0003] In an inspection device of a manufactured product, etc., when an object of the same shape is photographed, it can be judged that dirt or foreign matter has been mixed in if there is a part with a different mapping method by comparing it with a previously ca...

AI Technical Summary

Problems solved by technology

This prior art has the following problems. It is necessary to install a movable part such as a lens or a diaphragm, thereby increasing the cost, and since the invariable part between a plurality of images is judged to be dirt, it is possible to use the background that is not dirt as dirt. Unchanged part is judged as dirt

Images