The Internal Thread Detection System By Image Processing Based On DSP

This thesis is to study a kind of image processing based on DSP for fastening nut eligibility detection scheme. Detection system mainly includes the internal thread of screw nut image acquisition, transmission, and image processing and image feature parameter extraction. By constructing platform for optical image acquisition of the internal thread image, based on the DSP image processing platform internal thread through image processing and feature extraction, and the characteristics of the internal thread parameters as the basis of judging nut eligibility, the judgment and detection of nut eligibility are completed.This thesis is respectively about two aspects of hardware and software in the field of image processing system. By using the design of the optical image acquisition platform for high quality internal thread images and the modulus of conversion to image processing system, simplifying the image processing algorithms and improve system processing speed, the main image processing system is studied in the hardware composition and software realization of image processing algorithms, including image preprocessing, edge detection, tilt correction, positioning and segmentation, and image thinning. The image is converted into easy to detect single pixel edge information, through internal thread feature extracting, get the parameters such as diameter, tooth high, pitch of thread in the screw nut.On the basis of the CCS software development environment, this thesis designs a digital image processing system based high speed digital signal processor DSP, which through the image processing method can quickly and accurately complete automotive fastening nut test and eliminate the unqualified products. Compared with the traditional direct contact of artificial detection method, the detection system is more stable and fast, and this provides a new train of thought and method for automotive fastening nut rapid industrial detection.