Abstract: An image of a part is used to judge whether it is good or no good. A predetermined intensity, edge or similar feature image is extracted from the inspected image and is quantized. The quantized feature image is obtained from a training image in advance, and the feature is subjected to a generalized Hough transform operation which refers to the weight of each feature point for each feature value in the image and vote the feature point weight. The similarity is obtained from the results of the transform operation, and a good/no good judgement is made depending upon whether the similarity is above or below a threshold value.

Abstract: A color picture is separated into three color component pictures, which are divided respectively by first, second and third block coding means into blocks and encoded. In the first block coding means, one of the color component pictures is encoded, for each block, into gray level codes representing gray level components in the block and a resolution code representing the distribution of the gray level codes in the block. In this case, when the gray level in the block undergoes a little change, an average gray level can be used as the gray level code and the resolution code can be omitted. The second and third block coding means may be identical in construction with the first block coding means. The resolution code of the block obtained by the first block coding means can also be employed in the coding of the corresponding blocks by the second and third block coding means.

Abstract: A frame of a gray-scaled picture is divided into blocks, in each of which the mean luminance of picture elements is used as a threshold value for comparison with each picture element signal to classify it into 0 or 1 according to its magnitude to provide a resolution component. From the resolution component and each picture element signal in the block are calculated two gray components; the resolution component and the two gray components are used as coded outputs for each block.

Abstract: A frame of a gray-scaled picture is divided into blocks, in each of which the mean luminance of picture elements is used as a threshold value for comparison with each picture element signal to classify it into 0 or 1 according to its magnitude to provide a resolution component. From the resolution component and each picture element signal in the block are calculated two gray components; the resolution component and the two gray components are used as coded outputs for each block. The size of each block is changed in accordance with the property of the picture in the portion corresponding to the block.