Tool for fast and robust recognition of polymorphic objects
CVB Polimago is wellsuited to any application where the target is variable, this includes industrial applications were defects might be difficult to closelyspecify.
The key features of CVB Polimago are small training data requirements and low computing requirements. This enables applications that would otherwise not be feasible including many industrial applications. CVB Polimago also handles more of the complexity of the learning process than most Convolutional Neural Network (CNN) tools, bringing machine learning capabilities to more users.
In recent years deep learning and machine learning have become hot topics in machine vision. Many of these have been based on neural networks, but these have intrinsic drawbacks such as a requirement for a large number of training images and a heavy processing load, especially during training.
CVB Polimago has the advantages of other machine learning tools, such as the ability to search for and classify variable targets, but avoids some of the drawbacks. Typically CVB Polimago needs tens of images per class, compared to five hundred to a thousand for a neural network tool. Whereas most neural networks rely on GPUs to be fast enough for industrial use, CVB Polimago runs on a standard CPU, and often runs faster than neural networks. This makes it wellsuited to many machine vision environments, where training and execution time and the numb er of training images are constraints. In tests, CVB Polimago returns similar levels of accuracy to neural network approaches.
CVB Polimago applications divide into two types ‒ search and classification. Search applications include finding variable defects that are not easily found by classical methods. Classification might follow a search step, to allow a target to be found and then classified as a specific type. One example of this is complex OCR, such as stamped characters or handwriting.
A niche usage of CVB Polimago is 3D pose estimation. For a planar target it is possible to learn the effects of perspective as well as rotation, scaling and variability. This means that CVB Polimago can return the x,y position, the scale and also the 3D axial angles (alpha, beta, gamma) ‒ resulting in description of the position and 3D direction that the object is facing.
CVB Polimago applications are many and varied ‒ they include the search (and tracking) of organic or continuouslyvariable objects. This makes it a good choice for the food industry.
The variability of a target's appearance might also be due to illumination, so CVB Polimago can be a good choice for outdoor applications or those with uncontrolled lighting, such as agriculture and traffic applications.
CVB Polimago classification applications can be similar to the search applications, except that a decision is made between multiple trained classes. OCR, gender classification, vehicletype recognition and defect classification are all examples of this.