Real video time size selection mode

Claim 1

A method for real video time size selection for objects represented as video images which are to be filtered out so as to be classified as to their shape, comprising: recording the video image with a scanning recording beam; digitalizing the video image; providing at least one selected mask representing a selected object image shape; during each pixel clock period of the video image, comparing a k×L pixel section of the digitalized image, pixel for pixel, with the selected mask, using comparison elements in the form of random access memories programmed as look-up tables, while displacing the selected mask over the video image in synchronism with the recording beam, with said step of comparing including separately comparing k packets of L pixels at a time with the respective pixel values of the mask, with the product of k×L being the total number of pixels to be compared during each pixel clock period, assigning a desired comparison value to each pixel packet, determining the deviation of the actual comparison value from the desired comparison value for each pixel packet and providing a respective deviation value corresponding to each deviation, adding the resulting deviation values for the k packets, comparing the addition result with a threshold value, and checking each pixel packet for validity; and, indicating detection of the object on the recorded video image when, during said step of comparing a k×L pixel section, all of said k pixel packets which are being compared at the same time are determined to be valid and said threshold value is exceeded.

Claim 2

A method for real video time size selection for objects represented as video images which are to be filtered out so as to be classified as to their shape, comprising: recording the video image with a scanning recording beam; digitalizing the video image; providing at least one selected mask representing a selected object image shape; during each pixel clock period of the video image, comparing a k×L pixel section of the digitalized image, pixel for pixel, with the selected mask, using comparison elements in the form of random access memories programmed as look-up tables, while displacing the selected mask over the video image in synchronism with the recording beam, with said step of comparing including separately comparing k packets of L pixels at a time with the respective pixel values of the mask, with the product of k×L being the total number of pixels to be compared during each pixel clock period, assigning a desired comparison value to each pixel packet, determining the deviation of the actual comparison value from the desired comparison value for each pixel packet and providing a respective deviation value corresponding to each deviation, adding the resulting deviation values for the k packets, comparing the addition result with a threshold value, and checking each pixel packet for validity; and, indicating detection of the object on the recorded video image when, during said step of comparing a k×L pixel section, all of said k pixel packets which are being compared at the same time are determined to be valid and said threshold value is exceeded.